Leilani Exbom is morbidly obese. She has been so for most of her life. This condition does not suit her, and she has not accepted it passively. Indeed, for the last ten years she has waged a virulent campaign against it. She has made repeated attempts to reduce her weight by dieting. Dieting, however, has availed her little improvement. She has also tried to reduce her weight by surgery. In 1979 she underwent what is known as a jejunoileal bypass and subsequently lost 110 pounds. The surgery was risky, however, and in 1986, as the threat of unfortunate side effects increased, it was undone. Exbom then regained the weight she had lost.

By Spring of 1987 Exbom's obesity had created some problems. Exbom had a backache, her muscles were sore, and her limbs felt heavy. She was often tense and keyed up, and then fatigued. Her self-esteem was low, and she was self-conscious around others. She had low energy and little interest in the normal activities of life. She felt lonely, even blue. As a cure for these problems her doctor recommended a gastroplasty, or gastric stapling operation, which purportedly would reduce her ability and desire to eat. This, in turn, purportedly would lead to her long sought weight loss and an improvement in her physical condition.

Gastroplasties are expensive. Exbom wanted one, but she did not want responsibility for the bill. She sought instead to have the Central States, Southeast and Southwest Areas Health and Welfare Fund pay for the proposed operation. Central States is a Taft-Hartley Trust governed by an eight member Board of Trustees. Four of the Trustees represent the International Brotherhood of Teamsters and the other four represent the management of businesses in the trucking industry. The Trustees administer the Trust in accordance with a Trust Agreement. The Agreement authorizes the Trustees to make rules and regulations necessary to administer the Trust, to establish plans to provide benefits to certain teamsters and their families, to be the final authority in disputes about benefits, and to interpret conclusively the terms of the Agreement and the plans.

One of the plans set up by Central States is the C-6 Medical Plan. The Plan pays for health-related expenses incurred by its members, one of whom was Exbom. The Plan does not pay for these expenses willy-nilly, however. There are limitations. Two of them are important in this case.

One is the cosmetic surgery limitation. The Plan will not pay "any charge incurred for any treatment or service connected with surgery performed for cosmetic purposes ... unless the treatment service is required as a result of accidental bodily injury." The Plan Document, which is the technical constitution for the Plan, defines "cosmetic" as "[r]econstructive or plastic surgery which is provided primarily to improve the physical appearance of a Covered individual and which restores appearance but does not correct or materially relieve a medical condition." The Document provides an example of cosmetic surgery: "Surgery for obesity, including gastric bypass, gastric stapling [gastroplasty], intestinal bypass, lipectomy, suction lipectomy, and any other surgical procedure, the purpose and result of which is simply to remove adipose tissue." The Plan Booklet, which is an explanation of the Plan in less technical language, recites similar language. The other limitation is the experimental surgery limitation. The Plan will not pay "any charges [for medical treatment] which are not uniformly and professionally endorsed by the general medical community as standard medical care, including [medical treatments] which are experimental in nature." The Document defines "experimental" as charges for medical treatment "not uniformly and professionally endorsed by the general medical community as standard medical care." The Booklet again recites similar language.

In May of 1987 Exbom submitted to Central States a request for benefits on a questionnaire entitled "Predetermination of Benefits for Surgery for Obesity." This form had been brought to Exbom's attention via a Central States newsletter. The newsletter informed her that "[i]f determined medically appropriate, Gastric Stapling is the only obesity surgery eligible for benefit coverage approval." It instructed her that the Predetermination for Benefits form had to be filled out prior to surgery to allow Central States a chance to determine if the proposed surgery was "medically appropriate."

Exbom's form was reviewed by three medical consultants for Central States. All three recommended that her request for coverage be denied. Their view was that the proposed surgery was both cosmetic and experimental in nature and therefore unpayable under the terms of the Plan. Central States adopted their recommendation and denied Exbom's coverage request.

She appealed. Central States has an appeal process as mandated by section 503 of the Employee Retirement Income Security Act of 1974, 29 U.S.C. Sec. 1133. The process consists of three steps. In the first step ("Level I") Central States's Health and Welfare Communications Department reviews the claimant's request for benefits and the relevant evidence, including medical opinions. As the subject matter of the appeal is inevitably medical in nature, the decisionmakers maintain medical consultants and often rely heavily on opinions provided by them. This is so not only in Level I, but throughout the appeal process. In the second step ("Level II") a Claim Appeal Committee reviews the Level I decision, the request, and the relevant evidence. In the third step ("Level III") Central States's Board of Trustees reviews the request, all of the evidence, and all of the prior decisions, and then makes a final, binding decision. At each stage in the appeal process the claimant may submit additional information that she wants the decisionmakers to consider. The claimant may also review the appeal file, so her submitted information may address the concerns communicated by the decisionmakers.

Exbom's appeal was denied at both Level I and Level II. She took it to Level III, where in November of 1987 the Board of Trustees considered her request for coverage. In so doing the Trustees reviewed a mass of information, including statements from several of Exbom's doctors. The gist of these statements was that Exbom was morbidly obese, that her morbid obesity had led to physical and psychological problems, that she was likely to remain obese without a gastroplasty, and that a gastroplasty was an "accepted mode of treatment in the United States." Also among the information were the recommendations of Central States's medical consultants. One of these consultants, Dr. Buckingham, had reviewed Exbom's case three times (each time as new information was submitted). Dr. Buckingham was of the opinion that the medical benefits of gastroplasties were unknown. The literature on the procedure had failed to establish that it could cure obesity-related problems, including the problems Exbom purportedly was suffering from. Moreover, Dr. Buckingham felt that the medical risks of gastroplasties were unknown. He noted that weight-loss surgery often resulted in later complications, and pointed out that Exbom's weight-loss operation in 1979 had to be revised in 1986 for this very reason. Due to these uncertainties and on the basis of all the evidence, Dr. Buckingham believed that Exbom's proposed surgery was experimental in nature.

He also felt that it was cosmetic. He noted that Exbom was then 48 years old and suffering from a variety of obesity-related health problems. He opined that "the proposed surgery does not cure diseases related to obesity" and, therefore, the surgery was not medically necessary. The doctor admitted that in the past he had recommended payment for other gastric stapling procedures, but said that he had recommended those surgeries only for obese patients who were young and who suffered no obesity-related medical problems. He recommended the surgery in the hope that it would prevent those problems from occurring at a later date. It was his opinion, however, that the surgery was not medically helpful (although it could be aesthetically pleasing) once problems had developed. And in this case problems had developed.

In light of Dr. Buckingham's opinion, and after considering all other documents, evidence, and prior opinions and interpreting the terms of the C-6 Plan, the Trustees denied Exbom's Level III appeal. They found that the surgery she proposed was cosmetic and experimental in nature. By letter dated November 25, 1987, she was informed of their decision.

In June of 1988 Exbom filed suit in federal court against Central States. She brought an action that sought to have the Trustees' denial of her benefits overturned pursuant to ERISA Sec. 502(a)(1)(B), 29 U.S.C. Sec. 1132(a)(1)(B).* Section 1132(a)(1)(B) gives a member of a plan standing to sue "to recover benefits due to him under the terms of the plan, to enforce his rights under the terms of the plan, or to clarify his rights to future benefits under the terms of the plan." Exbom's suit progressed until it came before the district court on cross-motions for summary judgment. In an opinion entered March 21, 1989, the district court disposed of the action by granting Central States's motion. The court found no reason to upset the Trustees' decision; in its view they did not act arbitrarily or capriciously in denying Exbom's claim.

Exbom now argues that the district court erred, for two reasons. First, she believes that the district court used the wrong standard of review. Exbom admits that the district court used the standard of review held in a long line of Seventh Circuit cases to be proper in ERISA benefit denial reviews: the "arbitrary and capricious" standard. But she feels that a Supreme Court case handed down subsequent to the district court's opinion, Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 109 S. Ct. 948, 103 L. Ed. 2d 80 (1989), wiped out the basis for this standard and impliedly mandated one more strict. Second, she believes that regardless of the standard of review the Trustees' decision was arbitrary and capricious. She alleges that it was so in two ways: (1) it was arbitrary and capricious because the Trustees do not summarily reject requests for gastroplasty coverage, even though the Plan Document lists "gastric stapling" as an example of cosmetic surgery; and (2) it was arbitrary and capricious because the evidence presented to the Trustees is insufficient to support a finding that gastroplasty procedures are experimental.

We begin with the standard of review. The Supreme Court's case in Firestone provides some guidance. In that case the Court held that "a denial of benefits challenged under Sec. 1132(a)(1)(B) is to be reviewed under a de novo standard unless the benefit plan gives the administrator or fiduciary discretionary authority to determine eligibility for benefits or to construe the terms of the plan." Firestone, supra, at ----, 109 S. Ct. at 956. It then proceeded to employ the de novo standard in its review. But Firestone 's de novo standard is not for this case, and Exbom admits as much. Article IV, Section 17 of the Trust Agreement gives the Trustees power to construe the provisions of the Agreement and the Plan, and also states that any construction adopted by the Trustees in good faith is binding. Article V, Section 2 provides that all controversies over anything, including a denial of benefits, related to the Trust or Plan must be submitted to the Trustees and that their decision on the matter shall be binding. The effect of these provisions is to entrust the Trustees with discretionary authority to determine both Exbom's eligibility for benefits and the proper construction of the Plan's terms. This leaves the Trustees' decision under the "unless" clause of the Firestone standard, and outside the scope of de novo review. See Bali v. Blue Cross & Blue Shield Ass'n, 873 F.2d 1043, 1047 (7th Cir.1989).

But perhaps we can draw more from Firestone than its narrow holding. It is true, as Exbom insists, that Firestone was a step away from the dogmatic application of the "arbitrary and capricious" standard of review in all Sec. 1132(a)(1)(B) cases. Before Firestone many courts applied that standard even when the decision under review was not a discretionary one. They did so on the basis of an analogy drawn between ERISA and the Labor Management Relations Act ("LMRA"). Since the decisions of LMRA fiduciaries were reviewed under the arbitrary and capricious standard, decisions of ERISA fiduciaries were also reviewed under that standard. But the Court showed the analogy to the LMRA to be a false one, and thereby destroyed the basis for ERISA's arbitrary and capricious standard.

Because reference to the LMRA was faulty, the Court opined that "the wholesale importation of the arbitrary and capricious standard into ERISA is unwarranted." Firestone, supra, at ----, 109 S. Ct. at 953 (emphasis original). The Court stated that the common law of trusts, not the LMRA, was the proper source from which to draw ERISA legal standards. Id., at ----, 109 S.Ct. at 954. As the common law of trusts mandated a de novo review for the case before the Court, the Court held that such a review was mandated by ERISA. Id., at ----, 109 S.Ct. at 955. But the Court did not tell what standard should be used in situations like this, where de novo review is improper, other than to say that it should be "deferential." Id., at ----, 109 S.Ct. at 954. Reference was made in the opinion, however, to several possible standards. The Court's statement about an unwarranted wholesale importation into ERISA of the arbitrary and capricious standard suggests that in cases of discretion a partial importation of that standard may be warranted. See id., at ----, 109 S. Ct. at 953. See also Lakey v. Remington Arms Co., 874 F.2d 541, 544 (8th Cir.1989) (adopting arbitrary and capricious standard post-Firestone ). The Court's quotation of the Restatement (Second) of Trusts Sec. 187 (1959) suggests that an "abuse of discretion" standard may be proper. See Firestone, supra, at ----, 109 S. Ct. at 954. See also Boyd v. Trustees of the United Mine Workers Health & Retirement Funds, 873 F.2d 57, 59 (4th Cir.1989) (adopting abuse of discretion standard post-Firestone ); Lowry v. Bankers Life & Casualty Retirement Plan, 871 F.2d 522, 525 (5th Cir.) (same), cert denied, --- U.S. ----, 110 S. Ct. 152, 107 L. Ed. 2d 111 (1989). The Court also stated that a reasonableness standard may suffice: "A trustee may be given a power to construe disputed or doubtful terms, and in such circumstances the trustee's interpretation will not be disturbed if reasonable." Firestone, supra, at ----, 109 S. Ct. at 954. Finally, the Court hinted that a "good faith" standard may be proper in this case, since the Central States Trust Agreement speaks of complete discretion in the Trustees, subject only to a good faith restraint. See id., at ----, 109 S. Ct. at 954. See also Central States, S.E. & S.W. Areas Pension Fund v. Central Transport, Inc., 472 U.S. 559, 568, 105 S. Ct. 2833, 2839, 86 L. Ed. 2d 447 (1985).

Despite presenting this smorgasbord of possibilities, Firestone is clear in its message. At the heart of the case is an admonition to courts that they refer to the common law of trusts for guidance in ERISA decisions. Under the common law courts will not review the discretionary decisions of trustees and other fiduciaries de novo, but will look only for the trustee's abuse of its discretionary authority. Restatement (2d) of Trusts Sec. 187 (1959). What constitutes an abuse of discretion, however, depends in the first instance on how the trustee's discretion is defined. Where a trust agreement states that a trustee's discretionary decisions are binding if reasonable, an abuse of discretion occurs when the court cannot uphold the trustee's actions under a standard of "reasonableness." See Bali, 873 F.2d at 1047. Where, as here, the trustee's discretion is completely unrestrained by the language of the trust agreement or limited only by the requirement that decisions be taken in good faith, the court should review the trustee's actions under a more deferential standard. Courts facing a decision by a trustee vested with this highest level of discretion should review the trustee's decisions under a standard embodying the highest level of deference. That standard is the arbitrary and capricious standard. Egert v. Connecticut Gen. Life Ins. Co., 900 F.2d 1032, 1035 (7th Cir.1990).

The arbitrary and capricious standard holds that a trustee's decision shall not be overturned on a Sec. 1132(a)(1)(B) matter, absent special circumstances such as fraud or bad faith, if "it is possible to offer a reasoned explanation, based on the evidence, for a particular outcome." Pokratz v. Jones Dairy Farm, 771 F.2d 206, 209 (7th Cir.1985). "[A] court will not set aside the denial of a claim if the denial is based on a reasonable interpretation of the relevant plan documents." Shull v. State Machinery Co., Inc. Employees Profit Sharing Plan, 836 F.2d 306, 308 (7th Cir.1987). Nor will it do so where the trustee has based its decision " 'on a consideration of the relevant factors' " that encompass the " 'important aspect[s] of the problem' " before it. See Reilly v. Blue Cross & Blue Shield United of Wis., 846 F.2d 416, 420 (7th Cir.) (quoting Motor Vehicle Mfr's Ass'n of the U.S., Inc. v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 2866, 77 L. Ed. 2d 443 (1983)), cert. denied, --- U.S. ----, 109 S. Ct. 145, 102 L. Ed. 2d 117 (1988). If the trustee makes an informed judgment and articulates an explanation for it that is satisfactory in light of the relevant facts, i.e., one that makes a "rational connection" between the issue to be decided, the evidence in the case, the text under consideration, and the conclusion reached, then the trustee's decision is final. See Smart v. State Farm Ins. Co., 868 F.2d 929, 936 (7th Cir.1989). Cf. Van Boxel v. Journal Co. Employees' Pension Trust, 836 F.2d 1048, 1049-53 (7th Cir.1987) (examining arbitrary and capricious standard in detail).

Under this standard we cannot possibly overturn the decision of the Trustees. They were impartial judges. The issue before them was clear. They gathered all of the relevant information. They afforded Exbom exemplary process. They examined all of the evidence, and did so in depth. They were well advised by experts from the medical community. They formed a reasonable construction of the Plan's language. And they reached a judgment that was informed and supported by sound ratiocination. Their decision was not arbitrary, nor was it capricious. If anything, it was the opposite. It was well grounded in reason, well supported by the evidence, and well backed by the terms of the Plan. It would pass muster under any standard of deferential review.

This is so, Exbom's arguments to the contrary notwithstanding. Exbom asserts that in light of Dr. Buckingham's past recommendations to cover gastroplasty procedures, such procedures cannot be "cosmetic" under the terms of the Plan. Accordingly, the Trustees' rejection of her request on the "cosmetic" grounds must have been arbitrary. We agree that, despite being listed as an example of cosmetic surgery, gastric stapling in general is not always "cosmetic" in nature. The Plan makes clear that cosmetic surgery is surgery without medical benefits, and Dr. Buckingham has said that at one time he believed that gastroplasties could provide medical benefits for certain patients. Thus, at least in Dr. Buckingham's opinion, a gastric stapling procedure could avoid the cosmetic label in some cases. The fact that Central States does not summarily reject requests for gastroplasty coverage supports this reading of "cosmetic." Apparently, then, the Plan does not automatically rule out benefits for gastroplasties. But this does not make the converse true, that gastroplasties are always covered under the Plan. In some cases the procedure may be cosmetic; in others it may not. The evidence shows that the Trustees reasonably believed that a gastroplasty in this case would not "correct or materially relieve a medical condition" and, therefore, would do nothing more than "improve the physical appearance of a Covered individual." Hence, they opined that the surgery was cosmetic. The opinion was reasonable. It was well supported by the distinction made by Dr. Buckingham between those patients for whom gastric stapling coverage had been recommended and Exbom, for whom it had not. The Trustees' decision, therefore, was not arbitrary and capricious.

Exbom also asserts that the evidence cannot support the Trustees' determination that her proposed surgery was experimental in nature. She points to the evidence she submitted in which several doctors recommended the gastric stapling surgery. In her view, the recommendations of these doctors constitute a "general medical community" that "uniformly and professionally endorsed [gastroplasty] ... as standard medical care." She also points to the fact that Dr. Buckingham has recommended coverage for gastroplasties involving young, healthy, obese persons in the past. She argues that a gastroplasty is either experimental or it is not, and that this recommendation history shows the Trustees' efforts to have it both ways.

We have several problems with Exbom's argument, first of which is her claim about the sufficiency of the evidence. True it is that the Trustees had before them the recommendations of Exbom's doctors. But they also had before them the reasoned opinions of Dr. Buckingham and others that the surgery proposed for Exbom was experimental. These opinions are sufficient to support the Trustees' decision. Exbom is really asking the Court to reweigh the evidence on this issue, and that we will not do. Moreover, the recommendation of Exbom's doctors should not be confused with an opinion that the surgery is not experimental under the Plan. A gastroplasty may not be experimental when performed to cure obesity, but it may be when performed to cure already prevalent medical problems caused by obesity. It was the opinion of Central States's medical consultants and the Trustees that the gastroplasty would not cure Exbom's medical problems, and may even exacerbate them. It was therefore experimental in the medical sense even though it may be a proven method for curing obesity, and it is the medical sense that is important under the terms of the Plan. We also have a problem with Exbom's characterization of Dr. Buckingham's recommendation history. The fact that Dr. Buckingham recommended payment for gastroplasties in the past only means that, in his opinion, the surgery was not experimental in those particular cases. He has explained how those cases differ from Exbom's and why they cannot be compared. We simply cannot craft a general rule from those limited cases, as Exbom would have us do.

Exbom's arguments have failed to move us from our conclusion that the Trustees' decision denying her benefits was not arbitrary and capricious. It seems clear to us that the Trustees did not abuse the discretion given them by the Trust Agreement, and that their decision must stand.

For the foregoing reasons, the judgment of the district court is AFFIRMED.