Opinion by Justice CLEO E. POWELL.

Steven K. Funkhouser ("Funkhouser") brought a products liability action against Ford Motor Company and Obaugh Ford, Inc. (collectively "Ford") after his daughter, Emily, died from severe burns she suffered as a result of a fire in her family's Ford Windstar van. In this appeal, we consider whether the circuit court erred in excluding evidence of seven other Ford Windstar fires and in ruling that Funkhouser's expert witnesses could not rely on the excluded evidence. For the reasons that follow, we affirm the circuit court's rulings.

I. Facts and Proceedings Below

On May 4, 2006, Emily and Evan Funkhouser, three-year-old twins, were playing in their parents' 2001 Ford Windstar. The engine was off and the keys were not in the ignition. At some point, a fire erupted in the passenger compartment of the van. Emily suffered significant third-degree burns and, as a result, died later that afternoon.

In August 2007, Funkhouser, as administrator of Emily's estate, filed a wrongful death action against Ford alleging "a design defect in a particular electrical connector behind the dashboard of the Ford Windstar van that caused it to ignite." After Ford was granted a motion in limine to exclude evidence of other Windstar fires, Funkhouser took a voluntary nonsuit.

In January 2010, Funkhouser again filed a wrongful death action against Ford, alleging negligence and breach of implied warranty. Funkhouser's action was based on the theory that Ford failed to adequately warn consumers about the fire hazards existing in Windstar vans when they are parked with the engine off and no key in the ignition.

A. Cause and Origin of the Funkhouser Fire

According to Funkhouser's designated expert, Michael J. Schulz ("Schulz"), the origin of the fire in Funkhouser's vehicle "was located within the vehicle's instrument panel area with the key in an off position" and "[a]lthough there are multiple options within the instrument panel and surrounding area that could explain the electrical fire, the most likely origin point of the fire was in the lower portion of the panel in the vicinity of the wiring harness, cigarette lighter and the controls for the heating and AC system." Schulz went on to explain that the fire was caused

Relying on documents produced by Ford during the pendency of Funkhouser's first action,2 Schulz opined that

B. Evidence of Other Windstar Van Fires

As evidence that Ford knew or had reason to know that there was a danger of key-off dashboard electrical fires in its Windstar vans, Funkhouser sought to introduce evidence of other Ford Windstar fires that occurred prior to the Funkhouser fire. Relying on the reports produced by Ford, Funkhouser identified seven other Windstar fires that he asserted put Ford on notice of the danger of key-off dashboard electrical fires in its Windstar vans.

1. Mulkey Fire

In 2003, a 1999 Ford Windstar LX caught fire while the van was parked and not in operation. An investigation determined that the fire "began at the interior driver and center dash area" and "was caused by a failure of the wiring harness conductors and/or adjacent components located at the interior center and driver side dash area." However, "[t]he exact mechanism of failure [was] unknown," due to the "heavy degradation of the components and wiring conductors at the interior dash area."

2. Tirone Fire

In 2004, a 2003 Ford Windstar SEL caught fire when the van was parked and not in operation. An investigation determined that the fire began in the "dashboard area from the center section over to the left side" and was "accidental electrical in nature" as indicated by the "heavily shorted and beaded" wiring harness in the dashboard. The investigator noted that the vehicle had experienced problems with its electrical system prior to the fire.

3. Arencibia Fire

In 2004, a 1999 Ford Windstar LX caught fire while it was parked in a dealership service department repair shop and not in operation. There was no investigation into the cause or origin of the fire, only reports that it originated underneath the dashboard.

4. Bryan Fire

In 2002, a 1999 Ford Windstar caught fire while parked in a parking lot. According to the owner, the vehicle "blew up ... due to an electrical concern." However, the cause and origin of the fire was undetermined, as there was no investigation into the matter. Ford did note that the owner or his girlfriend indicated that fire may have been caused by arson.

5. Carf Fire

In 2000, a 1999 Ford Windstar LX caught fire while parked in the owner's garage and not in operation. An investigation determined that the fire originated "in the area of the dashboard" and was "caused by an electrical malfunction within the dashboard." However, "[d]ue to the complete destruction of the interior of the ... vehicle by the fire, a more complete precise cause could not be isolated." The investigator further noted that electrical problems preceding the fire indicated "a serious electrical malfunction which was causing a large current drain."

6. Pell Fire

In 2003, a 2002 Ford Windstar LX caught fire while the van was parked and not in operation. A fire department report states that the fire was "up under [the] glove box." According to the owner of the vehicle, investigators from the fire department, police department and the insurance company determined that the fire resulted from an electrical malfunction.

7. Roth Fire

In 2002, a 1999 Ford Windstar LX caught fire while the van was parked and not in operation. The initial investigation by the Fire Marshall's office determined that the origin of the fire was "located in the engine compartment" and that "a mechanical malfunction could not be ruled out as a possible cause of [the] fire." A subsequent investigation by a forensic automotive engineering firm determined that the fire originated "beneath the left end of the instrument panel and behind the instrument cluster" and resulted from "an electrical abnormality localized to the wiring harness of the instrument cluster electronic circuit board."

C. Trial Proceedings

Following discovery, Ford filed a motion in limine asking the court to reconfirm its ruling from the previous action excluding evidence of the other Windstar fires. Ford argued that Funkhouser could not prove that the causes of the seven other fires were substantially similar to the cause of the Funkhouser fire. After a hearing on the matter, the court issued a letter opinion granting the motion in limine. The court determined that, because "[t]he exact defect is not known in the Funkhouser fire ... it is not fair to Ford to say it is the `same or similar defect and danger'" as those in the other seven fires. The circuit court noted that none of Funkhouser's experts stated in their depositions that the other seven fires occurred under substantially similar circumstances or had substantially similar causes as the fire at issue. The circuit court concluded by stating:

Funkhouser wrote to the court, requesting clarification as to whether the court's ruling precluded Funkhouser's expert witnesses from relying on the evidence of the other seven fires in forming their opinion as to whether Ford knew or had reason to know of the dangerous condition. On February 22, 2011, the circuit court entered an order memorializing the findings contained in its letter opinion and further holding that evidence of the prior fires was inadmissible "including as a predicate for the testimony of [Funkhouser's] expert witnesses."

In its Final Order on this matter, the circuit court explained that, in granting Ford's motion in limine it

Funkhouser appeals.

II. Analysis

On appeal, Funkhouser argues that the circuit court erred in excluding evidence of the seven other Windstar fires. In the alternative, Funkhouser contends that, even if evidence of the other Windstar fires was inadmissible, the circuit court erred in ruling that Funkhouser's experts could not rely on those other fires as a basis for their opinions.

A. Evidence of Other Fires

The issue before this Court is whether the circuit court erred in its determination that the proffered evidence of fires in seven other Windstar vans was inadmissible to establish that Ford had notice and actual knowledge of a defective condition. The issue relates to the admissibility of the evidence, therefore the proper standard of review is abuse of discretion. Midkiff v. Commonwealth, 280 Va. 216, 219, 694 S.E.2d 576, 578 (2010) (citing Coe v. Commonwealth, 231 Va. 83, 87, 340 S.E.2d 820, 823 (1986)).

In his amended complaint, Funkhouser proceeded on a theory that Ford failed to warn users of a known fire hazard in its Windstar vans. It is well established that, "[a] manufacturer is not an insurer of its product's safety, and a manufacturer has a duty to warn only if it knows or has reason to know that its product is dangerous." Owens-Corning Fiberglas Corp. v. Watson, 243 Va. 128, 134, 413 S.E.2d 630, 634 (1992). Thus, in bringing a failure to warn claim, a party must prove that the manufacturer:

Featherall v. Firestone Tire & Rubber Co., 219 Va. 949, 962, 252 S.E.2d 358, 366 (1979) (quoting Restatement (Second) of Torts § 388 (1965)).

To establish that a manufacturer knows or has reason to know of the danger in a duty to warn case, a plaintiff may present evidence of similar incidents, provided the prior incidents occurred "`under substantially the same circumstances, and had been caused by the same or similar defects and dangers as those in issue.'" Roll `R' Way Rinks, Inc. v. Smith, 218 Va. 321, 325, 237 S.E.2d 157, 160 (1977) (quoting Spurlin v. Richardson, 203 Va. 984, 989, 128 S.E.2d 273, 277 (1962)). "This rule springs from the lessons of human experience that similar causes can be expected to produce similar effects." Id.

In the present case, all Funkhouser can show is that the incidents occurred under substantially the same circumstances; he cannot show that the fires were caused by the same or similar defects. Indeed, Funkhouser implicitly concedes this fact, as he amended his initial complaint from a design defect claim to a failure to warn claim because he realized that he could not definitively prove the specific defect that caused the fire. Similarly, he cannot prove what defect, if any, caused the fires in the other vehicles. Therefore, the circuit court did not err in excluding the evidence of the other seven fires.

Funkhouser, however, asserts that this requirement actually results in an evidentiary threshold that is higher than what is required to prove the merits of his claims. Funkhouser notes that liability under a failure to warn claim does not require a showing of any defect, only a showing that the manufacturer "knows or has reason to know that its product is dangerous." Owens-Corning, 243 Va. at 134, 413 S.E.2d at 634. Thus, Funkhouser advocates that a relaxed substantial similarity test, where the terms "defects" and "dangers" are interchangeable, is necessary in failure to warn cases.

General Motors Corp. v. Lupica, 237 Va. 516, 521, 379 S.E.2d 311, 314 (1989) (emphasis added) (quoting Spurlin, 203 Va. at 989, 128 S.E.2d at 277). Thus, the substantial similarity test consists of two prongs: (1) the substantially same circumstances prong and (2) the causation prong. Removal of the defect requirement from the causation prong would allow a plaintiff to attribute notice and actual knowledge to a manufacturer based on the mere existence of a generalized danger; there would be no requirement for the danger to be attributable to the manufacturer in any way. This Court has previously rejected such generalized liability, recognizing that "[a] manufacturer is not an insurer of its product's safety." Owens-Corning, 243 Va. at 134, 413 S.E.2d at 634; see also Jones v. Ford Motor Co., 263 Va. 237, 254, 559 S.E.2d 592, 601 (2002) (requiring proof that the cause of the dangerous condition was a defect so as to attribute knowledge of that condition to the manufacturer); Lupica, 237 Va. at 522, 379 S.E.2d at 315 (rejecting evidence of a generalized dangerous condition that was not shown to have resulted from a defect in the manufacturer's product).3

Indeed, Funkhouser's relaxed substantial similarity test would undermine the entire causation prong of the test. By advocating the elimination of the requirement of similar defects from the test, Funkhouser is asking this Court to invert the test and infer similar causes, i.e., defects, from the existence of similar effects, i.e., fires. This inversion simply does not work: although a faulty cigarette lighter may cause a key-off dashboard electrical fire, not all key-off dashboard electrical fires are caused by a faulty cigarette lighter. Whether it is a products liability claim or a failure to warn claim, our jurisprudence establishes that the evidentiary test governing the admissibility of evidence relating to prior incidents must be strictly adhered to. To hold otherwise would allow a plaintiff to establish that a manufacturer knows or has reason to know of a danger based on prior incidents that were not attributable to that manufacturer. Therefore, we reject Funkhouser's argument that the court should adopt a relaxed substantial similarity test for the admissibility of prior incident evidence in failure to warn causes of action.

In order for the proffered evidence to be admissible to show that Ford knew or had reason to know that there was a danger of key-off dashboard electrical fires, Funkhouser must demonstrate that the other seven Windstar fires were caused by the same or similar defect. This does not mean that Funkhouser must identify the specific defect that caused the fire in his van or the other seven vans.4 As our cases have demonstrated, there are two avenues for a plaintiff in Virginia to establish substantial similarity in a failure to warn claim against a manufacturer: (1) through identification of the accident's cause, which must be attributable to the manufacturer, or (2) through the elimination of other potential causes that are not attributable to the manufacturer.5 See Jones, 263 Va. at 256-57, 559 S.E.2d at 602. In this case, Funkhouser can neither identify the cause of the fires nor rule out all other possible causes of the fires, including post-manufacture repairs, replacement parts in the dash, arson, or misuse. For this reason we hold that the evidence fails the substantial similarity test.

B. Expert Testimony

Funkhouser also argues the circuit court erred in ruling that his experts could not rely upon the evidence of the prior fires in their testimony regarding how a reasonable automobile manufacturer would react to those prior fires. Funkhouser relies upon Code § 8.01-401.1, which allows an expert to express an opinion or draw inferences from sources that are not allowed into evidence.

In pertinent part, Code § 8.01-401.1 states:

However, the fact that an expert witness may rely upon otherwise inadmissible evidence in forming an opinion, "does not... relieve the court from its responsibility, when proper objection is made, to determine whether the factors required to be included in formulating the opinion were actually utilized." Swiney v. Overby, 237 Va. 231, 233, 377 S.E.2d 372, 374 (1989). "If all the factors are not utilized, the court should exclude the opinion evidence." Id.

In examining such factors, this Court has recognized that, "[t]he results of experiments are not admissible in evidence unless the tests were made under conditions which were the same or substantially similar in essential particulars to those existing at the time of the accident." Featherall, 219 Va. at 959, 252 S.E.2d at 365. We have further held that an expert cannot offer opinion testimony based on such experiments because there are "`too many missing variables' to permit [an] expert to give his opinion." Mary Washington Hospital, Inc. v. Gibson, 228 Va. 95, 99, 319 S.E.2d 741, 743 (1984) (quoting Thorpe v. Commonwealth, 223 Va. 609, 614, 292 S.E.2d 323, 326 (1982)); see also Tittsworth v. Robinson, 252 Va. 151, 154, 475 S.E.2d 261, 263 (1996) ("[Expert] testimony cannot be speculative or founded upon assumptions that have an insufficient factual basis").

Along these same lines, we today hold that an expert cannot offer opinion testimony based on evidence that fails the substantial similarity test. To hold otherwise would be to allow an expert to offer an opinion based on speculative or otherwise irrelevant evidence. In the present case, none of the seven prior fires were the same or substantially similar to the Funkhouser fire. Thus, any expert testimony would necessarily be based on assumptions that have an insufficient factual basis. As such, there were too many missing variables to permit expert testimony based on those fires. Accordingly, the trial court did not err in precluding Funkhouser's experts from relying on the evidence of the seven other Windstar fires as a basis for their opinions.

III. CONCLUSION

For the foregoing reasons, we will affirm the ruling of the circuit court.

Affirmed.

Justice McCLANAHAN, with whom Justice LEMONS and Justice MIMS join, concurring in part and dissenting in part.

Justice McCLANAHAN, with whom Justice LEMONS and Justice MIMS join, concurring in part and dissenting in part.

In my view, the circuit court applied incorrect legal principles in ruling on the admission of evidence of prior Ford Windstar fires and in prohibiting the reliance upon such fires by Funkhouser's experts. The majority justifies the circuit court's rulings by crafting a new standard for admission of similar occurrences proof, incorporating this new standard into the elements of a cause of action for failure to warn, and creating its own exception to Code § 8.01-401.1.

A. Admission of Evidence Concerning Other Ford Windstar Fires

Because the admissibility of evidence depends on the plaintiff's theory of the case, the admissibility of the prior Ford Windstar fires must be determined in the context of Funkhouser's claim that Ford failed to warn of the danger of key-off electrical dashboard fires. See Breeden v. Roberts, 258 Va. 411, 416, 518 S.E.2d 834, 837 (1999) (evidence is relevant if "it tends to establish a party's claim or defense or adds force and strength to other evidence bearing upon an issue in the case"). Under Funkhouser's theory that Ford failed to warn of the danger of key-off electrical dashboard fires, Funkhouser must prove Ford (a) knew or had reason to know that the Funkhouser minivan was or was likely to be dangerous for the use for which it was supplied to Funkhouser, (b) had no reason to believe that Funkhouser would realize the minivan's dangerous condition, and (c) failed to exercise reasonable care to inform Funkhouser of the minivan's dangerous condition or the facts which make it likely to be dangerous. Featherall v. Firestone Tire & Rubber Co., 219 Va. 949, 962, 252 S.E.2d 358, 366 (1979) (applying Restatement (Second) of Torts § 388 (1965)). "A product is unreasonably dangerous if it is defective in assembly or manufacture, unreasonably dangerous in design, or unaccompanied by adequate warnings concerning its hazardous properties." Morgen Industries, Inc. v. Vaughan, 252 Va. 60, 65, 471 S.E.2d 489, 492 (1996). In this case, Funkhouser asserts the Windstar minivan supplied to it by Ford was unreasonably dangerous because it was unaccompanied by adequate warnings concerning the potential for key-off electrical dashboard fires.

We have traditionally permitted a plaintiff to prove notice of a dangerous condition through evidence of another similar incident or occurrence "`provided the prior incident occurred under substantially the same circumstances'" and was "`caused by the same or similar defects and dangers as those in issue.'" Jones v. Ford Motor Co., 263 Va. 237, 255, 559 S.E.2d 592, 601 (2002) (quoting Ford Motor Co. v. Phelps, 239 Va. 272, 276-77, 389 S.E.2d 454, 457 (1990) (quoting General Motors Corp. v. Lupica, 237 Va. 516, 521, 379 S.E.2d 311, 314 (1989))); see also Owens-Corning Fiberglas Corp. v. Watson, 243 Va. 128, 137, 413 S.E.2d 630, 635 (1992); Roll `R' Way Rinks, Inc. v. Smith, 218 Va. 321, 325, 237 S.E.2d 157, 160 (1977). Such evidence may only be admitted to prove notice and actual knowledge by the defendant of the dangerous condition, not to provide corroboration of the existence of such condition. Jones, 263 Va. at 255, 559 S.E.2d at 601. Thus, upon a timely request, a defendant will be entitled to a cautionary instruction informing the jury of this limited purpose. Roll `R' Way Rinks, 218 Va. at 327, 237 S.E.2d at 161.1

Applying these principles, the evidence in the record establishes the Mulkey, Tirone, Carf, and Roth fires "occurred under substantially the same circumstances" and were "caused by the same or similar defects and dangers as those in [the Funkhouser fire]." Jones, 263 Va. at 255, 559 S.E.2d at 601. All four fires occurred when the vans were parked, not in operation, and with no key in the ignition. The cause and origin of each of the fires was professionally investigated and determined to be electrical in nature, to have originated in the dashboard area of the vans, and to have been caused by the failure of electrical wiring or components within the dashboard area. The information regarding these fires contains no evidence of arson, misuse or some external cause for the fires. Since Funkhouser claims that his minivan was unreasonably dangerous for its intended use due to the danger of key-off electrical dashboard fires, evidence of these four Windstar van fires is admissible to prove Ford had notice and actual knowledge of the danger of key-off electrical dashboard fires.2

With regard to the Arencibia, Bryan, and Pell fires, however, I agree the evidence regarding these fires does not sufficiently establish that they were caused by the same or similar defect and danger as that alleged in the Funkhouser fire. While these fires occurred when the vans were not in operation and with no key in the ignition, there is no evidence of any investigation into the cause or origin of these fires. Absent sufficient evidence that these fires were caused by the failure of electrical wiring or components in the dashboard area, the evidence does not show that they were caused by the same or similar defects and dangers as the Funkhouser fire.

In ruling that all seven fires were inadmissible, the circuit court erred, in the first place, by framing the issue before it as whether Ford should be charged with notice and knowledge of a defective condition requiring warning of that condition. In particular, the circuit court stated that "[t]he legal issue here is whether Ford should be charged with notice and actual knowledge of a defective condition requiring the warning of that defective condition." (Emphasis by court.) Ruling that the Funkhouser defect must be "identified with specificity to charge Ford with actual notice of that defect," the court concluded the required specificity was absent such that it was "not fair" to charge Ford with notice of a defective condition. The issue before the court, however, was whether the other Windstar fires occurred "under substantially the same circumstances" and were caused by "the same or similar defects and dangers" as those alleged in the Funkhouser fire. Jones, 263 Va. at 255, 559 S.E.2d at 601 (internal quotation marks and citation omitted). Whether the Funkhouser minivan is unreasonably dangerous and whether Ford knew or should have known of the unreasonably dangerous condition are essential elements of Funkhouser's failure to warn claim and were not proper issues for the court to resolve on Ford's motion to exclude evidence of the other Windstar van fires.

The circuit court also applied incorrect legal principles in finding that "there is not enough specificity noted in the seven fires to say what the defect was that Ford had to warn of or correct" because the circuit court required Funkhouser to provide a level of specificity not required for a failure to warn claim such as this under Virginia law.3 Funkhouser is asserting that the minivan was unreasonably dangerous due to the potential for key-off electrical dashboard fires, not due to a specific design or manufacturing defect. Thus, the issue presented by Ford's motion to exclude evidence of the other Windstar van fires was whether the other fires were caused by the failure of electrical wiring or components within the dashboard area. Funkhouser was not required to allege a specific mechanical defect to establish the similarity of the fires.

We have previously found evidence of prior similar occurrences admissible to prove notice of a dangerous condition in the context of a failure to warn case without proof of a design or manufacturing defect. In Owens-Corning, we held that evidence of a summary of 44 workers' compensation claims filed by installers of insulation materials alleging they acquired lung diseases caused by exposure to asbestos dust was admissible in an action alleging Owens-Corning failed to warn of the dangers associated with use of insulation products containing asbestos. 243 Va. at 137, 413 S.E.2d at 635-36. As we concluded, the summary of workers' compensation claims was admissible to prove that "Owens-Corning had notice that insulators were at risk of contracting lung diseases from the use of insulation products which contained asbestos." Id.

As our analysis in Owens-Corning indicates, in determining whether other occurrences are caused by the same or similar defects and dangers, the terms "defects" and "dangers" are necessarily interchangeable in the context of a failure to warn claim since liability is based on the manufacturer's duty to warn "if it knows or has reason to know that its product is dangerous." Id. at 134, 413 S.E.2d at 634. The "substantial similarity" test was satisfied in Owens-Corning because the insulators in the workers' compensation claims alleged "they acquired lung diseases caused by exposure to asbestos dust while using insulation products," which was the same or similar dangers claimed by plaintiff. Id. at 137, 413 S.E.2d at 636.

Although the majority acknowledges that Funkhouser's theory is that Ford negligently failed to warn of the potential for key-off electrical dashboard fires, not that it negligently designed or manufactured its minivan, the majority defends the circuit court's analysis by creating additional factors that must be satisfied for the admission of similar occurrences proof. In Spurlin v. Richardson, 203 Va. 984, 989, 128 S.E.2d 273, 277 (1962), when this Court first enunciated the similar occurrences test, we stated that such evidence is admissible when "those prior accidents or occurrences happened at substantially the same place and under substantially the same circumstances, and had been caused by the same or similar defects and dangers as those in issue, or by the acts of the same person." Id. (emphasis added). Under the majority's standard, in order for evidence of prior occurrences to be admissible, a plaintiff must now identify as the cause of the prior occurrences a defect attributable to the defendant. Thus, not only must plaintiff identify a specific defect, the defect must be attributable to the defendant, which in this case is the manufacturer. Although the majority does not specify whether a design or manufacturing defect must be identified, it must be in the nature of one or the other since it must be attributable to this defendant.4 Accordingly, the majority's new standard for admission of prior occurrences can only be satisfied if a plaintiff can prove a cause of action for design or manufacturing defect.

According to the majority, the requirement that a plaintiff establish the prior similar occurrences were caused by a defect attributable to defendant is necessary because otherwise a plaintiff could "establish that a manufacturer knows or has reason to know of a danger based on prior incidents that were not attributable to that manufacturer." The majority's concern is misplaced because a plaintiff is not required to establish that the product's dangerous condition is caused by a defect attributable to the defendant in order to succeed under a failure to warn theory. Furthermore, the requirement that prior incidents be caused by a defect attributable to the defendant has never been a prerequisite to their admission. See Spurlin, 203 Va. at 989, 128 S.E.2d at 277 (such evidence is admissible when "those prior accidents or occurrences happened at substantially the same place and under substantially the same circumstances, and had been caused by the same or similar defects and dangers as those in issue, or by the acts of the same person") (emphasis added). The majority is conflating a cause of action based on negligent manufacture or negligent design with a cause of action based on the negligent failure to warn by reformulating the similar occurrences standard to require proof of a specific design or manufacturing defect and incorporating that requirement into the elements of a failure to warn claim.

It is beyond dispute that "[a] manufacturer is not an insurer of its product's safety." Owens-Corning, 243 Va. at 134, 413 S.E.2d at 634. The cause of action for failure to warn is not a theory of strict liability. The plaintiff must prove that the product was dangerous, that the defendant knew or had reason to know of this dangerous condition, and that the defendant had no reason to believe the plaintiff would realize the dangerous condition. Furthermore, the plaintiff must prove that the defendant failed to exercise reasonable care to warn of the dangerous condition of the product it supplied to the plaintiff and that its failure to exercise such care caused plaintiff's damages. See Featherall, 219 Va. at 962, 252 S.E.2d at 366 (discussing elements of such claims as set forth by the Restatement (Second) of Torts § 388). The jury is instructed on these elements, and it is unnecessary for this Court to augment the currently existing elements of the cause of action for failure to warn to ensure the jury follows its instructions to hold the defendant liable only where it knows or has reason to know of the product's dangerous condition.

The flaw in the majority's analysis becomes evident when it is applied to a supplier other than a manufacturer. A failure to warn claim can be asserted against any supplier of a product, and the elements are the same regardless of whether the defendant is the manufacturer or another person that supplies the product.5 The cause of the dangerous condition is not an element of a failure to warn claim because the negligence for which the supplier is held liable is the failure to exercise reasonable care in warning of the product's dangerous condition. Yet, by requiring a plaintiff to prove notice to a supplier through similar occurrences caused by defects attributable to that supplier, a plaintiff could never prove a failure to warn claim against a supplier other than a manufacturer.6 And against a manufacturer, the plaintiff could only prove a failure to warn claim if he or she can also prove negligent design or negligent manufacture. Thus, under the majority's test for establishing notice of a dangerous condition, the viability of a failure to warn claim in Virginia is substantially limited, if not entirely extinguished.

Unlike the majority, I would hold that evidence of the Mulkey, Tirone, Carf, and Roth fires is admissible. However, because I would hold that, under the proper analysis, evidence of the Arencibia, Bryan, and Pell fires is inadmissible, I concur in the majority's holding to that extent.

B. Admission of Expert Testimony

Funkhouser's experts would testify "on what the industry standard would be in response to at least seven reports of unexplained, key-off fires." Additionally, Schulz is of the opinion that other similar occurrences "should have placed" Ford on notice that Ford's Windstar minivans manufactured between 1999 and 2003 were or were likely to be dangerous for the use for which they were sold because Ford knew or should have known that the electrical components in the instrument panel area of these vans had the potential to fail and result in a fire with the key in an "`off' position."

Pursuant to Code § 8.01-401.1, "any expert witness may give testimony and render an opinion or draw inferences from facts, circumstances or data made known to or perceived by such witness at or before the hearing or trial during which he is called upon to testify." Furthermore, "[t]he facts, circumstances or data relied upon by such witness in forming an opinion or drawing inferences, if of a type normally relied upon by others in the particular field of expertise in forming opinions and drawing inferences, need not be admissible in evidence." Id. However, this statute does not allow for introduction of otherwise inadmissible evidence during direct examination of an expert witness merely because the expert relied on such evidence in formulating an opinion. See Commonwealth v. Wynn, 277 Va. 92, 100, 671 S.E.2d 137, 141 (2009). Therefore, Funkhouser's expert witnesses may not testify about or refer to any inadmissible fires during their direct testimony at trial.

Although Funkhouser's expert witnesses may not make reference to inadmissible fires during their direct examination, Code § 8.01-401.1 expressly permits expert witnesses to rely upon inadmissible information in formulating their opinions if it is "of a type normally relied upon by others in the particular field of expertise in forming opinions and drawing inferences." Thus, the circuit court's ruling prohibiting Funkhouser's experts from relying upon information regarding the other Windstar van fires in formulating their opinions is inconsistent with the language of Code § 8.01-401.1. Of course, Ford would be entitled to cross-examine Funkhouser's experts at trial as to the basis for each opinion, including whether, in formulating such opinion, the expert relied on occurrences not shown to be substantially similar to the Funkhouser fire. See id. (providing that the expert may be "required to disclose the underlying facts or data on cross-examination").7

Although the majority recognizes that Code § 8.01-401.1 permits expert witnesses to rely upon inadmissible information in formulating their opinions if it is "of a type normally relied upon by others in the particular field of expertise in forming opinions and drawing inferences," it upholds the circuit court's ruling prohibiting any reliance upon the prior fires by Funkhouser's experts by creating an exception that would prohibit experts from relying on evidence "that fails the substantial similarity test." Despite the well-established doctrine that "[c]ourts cannot `add language to the statute the General Assembly has not seen fit to include,'" the majority has amended Code § 8.01-401.1 to provide that information relied upon by an expert witness need not be admissible in evidence unless it is inadmissible because it fails the substantial similarity test. Jackson v. Fidelity & Deposit Co., 269 Va. 303, 313, 608 S.E.2d 901, 906 (2005) (quoting Holsapple v. Commonwealth, 266 Va. 593, 599, 587 S.E.2d 561, 564-65 (2003)).

Because I would apply Code § 8.01-401.1 as written, I dissent from the majority's holding that the circuit court did not err in precluding Funkhouser's experts from relying on information regarding the other Windstar fires in formulating their opinions. Therefore, while I conclude that evidence of the Arencibia, Bryan, and Pell fires is inadmissible, I would hold that Funkhouser's experts may rely upon the information regarding those fires in formulating their opinions if such information is "of a type normally relied upon by others in the particular field of expertise in forming opinions and drawing inferences." Code § 8.01-401.1.