WISDOM, Circuit Judge:

This is a trade secrets case. We do not, however, have to decide the difficult question whether the plaintiff, Texas Urethane, Inc. (Urethane), had a trade secret to be misappropriated. The defendant, Seacrest Marine Corporation (Seacrest), did not appeal a determination by Judge Noel that Urethane possessed a trade secret. 1 Texas Urethane, Inc. v. Seacrest Marine Corp., S.D.Tex.1975, 403 F.Supp. 612. The secret *137 was a new combination of old elements to produce a new and useful result — a rigid urethane foam that could be used as the inner part of a “sandwich” boat. Later, in a separate trial, Judge Sterling held that Seacrest had not misappropriated the secret. 2 Urethane appeals from this decision, contending that the district judge was clearly erroneous in his findings.

Seacrest was a manufacturer of boats for sale to retail distributors. 3 A low density liquid polyurethane foam was injected into a preformed mold made of an inner and outer skin of Acrylonitrile Butadiene Styrene (ABS) plastic. Borg-Warner Corporation made the plastic. The foam was intended to provide flotation, structural strength, and enough to adhesion to hold the two ABS skins together. It had to fill voids from IV2 inches thick to more than 20 inches thick and flow a minimum of twelve feet from the stern of the boat to the bow. The foam had to be inexpensive yet durable, and withstand testing of 200 hours in temperature up to 140 degrees Fahrenheit. Seacrest devised fielduse tests to minimize the number of boats generating warranty claims.

During 1970 and through the end of the summer of 1971, more than 21 chemical compounding companies competed to develop a satisfactory polyurethane foam. They undertook this expense because they expected Seacrest to purchase all of its foam from the successful foam supplier.

Polyurethane foam has two basic components. The “A” component is a liquid chemical, polymeric isocyanate. The “B” component is a liquid chemical consisting of four types of ingredients: a polyol, a catalyst, a blowing agent, and a surfactant. The polyol reacts with the A component side of the formulation. The catalyst controls the speed of the reaction. The blowing agent helps the material spread through the preformed mold. The surfactant regulates the size of the cells or bubbles appearing in the solidified foam. The most difficult part of making a polyurethane formula is determining the type, number, and proportions of “B” components.

Texas Urethane was one of the companies competing to develop a successful polyurethane foam. Dr. Bill Davis, its president and chief researcher, spent a great deal of time and effort in trial and error laboratory experimentation in finding a suitable foam. Although in 1970 and 1971 he did not produce an acceptable foam, he came closer to doing so than his competitors did. Consequently, representatives of Borg-Warner suggested a concerted effort to improve the formula and to reduce the failure rate of the boats. Dr. Davis, Dr. Luke Widenfelt, a Borg-Warner urethane expert, and Dr. James Payne, a Borg-Warner engineering expert, tried to “fine tune” the Urethane formula during August 1971. Borg-Warner and Dr. Davis executed a secrecy agreement to prevent either from revealing each other’s foam formula information to other suppliers. Seacrest was not a party to the agreement, but it did know of its existence.

The joint effort improved the formula somewhat, and Seacrest continued to purchase Urethane’s foam. In early 1972, however, Seacrest stopped ordering from Urethane and began to purchase foam from Union Carbide. 4 Eldridge Randolph Gar-retson, Seacrest’s Chemical Systems Manager, supplied Carbide with a formula for production of the foam. Seacrest hired *138 James William Porter to be Chemical Systems Manager and to continue developing their formula. He developed twenty-two formulas before he found a satisfactory foam. Seacrest’s position is that the formulas it used were derived from its own efforts; that they contain different ingredients, different proportions, react differently, and perform differently from the formulas Dr. Davis asserts as his trade secret formulations.

On July 28, 1972, Urethane filed suit in the 122nd Judicial District Court of Galveston County, Texas, against Seacrest and Union Carbide, asserting that they had misappropriated its trade secret. The defendants removed the case to the United States District Court for the Southern District of Texas. 5 On January 31, 1975, Judge Noel granted Seacrest’s request for a separate trial on the issue whether Urethane had a trade secret. He found that it had and that “Plaintiff’s trade secret is protected against misappropriation by this particular Defendant”. 403 F.Supp. at 619. On May 21, 1976, Chief Judge Garza transferred the case to Judge Sterling. Judge Sterling found that Seacrest had not misappropriated the trade secret:

Texas Urethane, Inc., through Dr. Davis, kept his formulation secret from defendant Seacrest except to the extent it was revealed to Seacrest in the context of a confidential relationship. Plaintiff made every reasonable effort to maintain this secrecy and Seacrest had no privilege to use or disclose the formulation. It could possibly be inferred from these circumstances in conjunction with the similarity of the Union Carbide formulation that defendant misappropriated plaintiff’s trade secret. However, the similarities in the formulations used by Seacrest at various times ... do not compel such an inference and, indeed, the preponderance of the credible evidence as to the dissimilarities and how they came about compels a contrary conclusion.

II.

The substantive law of North Carolina governs this case. Judge Noel correctly pointed out that little North Carolina law on trade secrets exists; accordingly, we must look to common law principles. The law is unclear. As one authority concludes, “[i]t appears that the rationale for protecting the trade secret often depends upon the facts of the particular case”. R. Callmann, 2 The Law of Unfair Competition Trademarks and Monopolies § 51 at 344 (3d ed. 1968). The decisions do agree that “a trade secret will only be protected against certain types of unfair or tortious conduct and not merely because it happens to be a secret”. Id. at 346. And most courts agree that there are three basic elements of a trade secret misappropriation cause of action: “(1) proof of the existence of an idea not generally known and treated by the owner as a secret; (2) improper disclosure or use; (3) proof of loss, either present or future”. Note, Theft of Trade Secrets: The Need for a Statutory Solution, 120 U.Pa.L.Rev. 378, 381-82 (1971); accord, Note, Common-Law and Statutory Sanctions for Industrial Espionage — A Need for Revision, 52 Iowa L.Rev. 63, 66-68 (1966).

We are concerned only with Judge Sterling’s finding that Urethane did not make out the second element of a misappropriation action — improper disclosure or use. A trial court’s rulings on question of fact are reviewable only to determine whether they are clearly erroneous. See Kentucky Fried Chicken v. Diversified Packaging, 5 Cir. 1977, 549 F.2d 368, 377; Fed.R.Civ.P. 52. “A finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” United States v. United States Gypsum Co., 1948, 333 U.S. 364, 395, 68 S.Ct. 525, 542, 92 L.Ed. 746, 766; Texas Distrib. v. Local Union No. 100, 5 Cir. 1979, 598 F.2d 393, 402. *139 We cannot say that the findings in this case are clearly erroneous.

Dr. Davis contended that his formula was unique because of the number of ingredients in the “B” component (nine) and because two of the ingredients, ethylene glycol and fluorocarbon 113, are very rarely used. Dr. George Speranza, a chemist employed by Jefferson Chemical Company (a Urethane supplier), testifying for the plaintiff, agreed that ethylene glycol and fluorocarbon 113 “are not used very frequently”. He thought that other parts of the formula were not rare; in fact, “a formulation very similar to this without the ethylene glycol and without the fluorocarbon 113 and without the fluorocarbon 11 is one that we published back in 1968. . . . And I think Bill [Davis] started with that formulation”. He thought using nine components was “unusual” but qualified this by saying that most suppliers try to give out similar formulas and that “there are probably other people that have done business in urethanes from the other end where they have taken our formulations, they have added to it and they have nine formulations, but we do not try to make a formulation with nine components”. He admitted that some publications might suggest using ethylene glycol.

Seacrest contended that the formula it submitted to Borg-Warner was primarily developed by Garretson. He had not been educated in urethanes but had picked up considerable knowledge from his own research and from the companies which had competed for the urethane foam contract. Davis and many other urethane experts told him what ingredients would make a suitable foam formula. He mentioned two other companies which had informed him that they used fluorocarbon 113 in their formulas. Borg-Warner had suggested using fluorocarbon 113 in Urethane’s foam.

Widenfelt of Borg-Warner testified that he had suggested using that type of fluorocarbon. He made this suggestion before the secrecy agreement was signed. He added that the ingredient was not uncommon, just expensive. Widenfelt said that ethylene glycol was also not unusual; indeed, it was a logical ingredient. Borg-Warner and “many other people” used it.

Payne of Borg-Warner was not sure who had made the recommendation to use fluorocarbon 113, but “suspected” that Widen-felt had done so. He thought that one other foam supplier competing for the Sea-crest contract might have used it.

At the trial before Judge Sterling, Dr. Davis admitted that Widenfelt may have suggested using the fluorocarbon 113:

Q. [Widenfelt] states that you all discussed [using fluorocarbon 113] and he had to more or less persuade you to go ahead and use it, and it may very well have been because you tried it before and thought it was too expensive, I’m not arguing with you, but my point is at that meeting didn’t he, friendly,-wise [sic], try to convince you to use 113?

A. I would say yes.

Q. And [the W in your formula] must have stood for Widenfelt [s/c]?

A. Okay.

Q. And that is because he is the one that thought up 113 at that time?

A. That’s right.

Q. Isn’t that correct? He is the one that thought up 113 at that time?

A. All right. Let’s say he is, I’m not sure.

Because testimony differed about the uniqueness of portions of the formula, the trial judge could reasonably conclude that the Seacrest formula was different.

The evidence also indicated that the proportions of the Garretson formulas were different from those of the various Urethane formulas. Some ingredients differed only slightly in quantity; other differences were greater. Moreover, Seacrest added one ingredient — water. Only the first of the four Urethane formulas contained water, and it contained 6% times as much as Garretson’s first formula.

Dr. Porter worked to develop a totally successful foam formula. He used a num *140 ber of different sources in the development of his formula, relying in part on textbooks and technical journals. He worked from Formula Number 6, a Garretson formula: “I knew Formula Number Six wasn’t working, and I knew I had to do something, but it did provide places to start. And so, Formula Six, and because of the easy availability of the base materials from the salesman that was calling on us once or twice a week, anyhow was a good place to work from.” His latter formulas do not contain ethylene glycol. Urethane has never contended that Porter misappropriated its secret directly; instead, Urethane argues that he based his work on a Garretson formula, which was copied from Urethane’s formula.

Some testimony did suggest that Seacrest may have taken Urethane’s formula. For example, Garretson testified that Davis gave him a sheet of paper listing some of the ingredients. Davis denied this. Both Payne and Widenfelt testified that the handwriting on the sheet was Widenfelt’s, but they did not know how Garretson got it. The sheet listed three ingredients, only one of which, ethylene glycol, Urethane contends is unique.

We cannot say that Judge Sterling’s determination that no misappropriation occurred was clearly erroneous. Urethane never explained how Seacrest got the formula it supposedly stole; the plaintiff merely states that the defendant must have obtained it somehow. The testimony of several witnesses indicates that there was no “improper disclosure or use” of Urethane’s formula. Admittedly, this testimony was controverted by Dr. Davis and, to a lesser extent, by other witnesses. But we must give “due regard ... to the opportunity of the trial court to judge of the credibility of the witnesses”. Fed.R. Civ.P. 52(a). “Credibility choices and the resolution of conflicting testimony are for the District Court, subject only to the clearly erroneous rule, Rule 52(a).” Hodgson v. H. Morgan Daniel Seafoods, Inc., 5 Cir. 1970, 433 F.2d 918, 920.

The judgment is AFFIRMED.

. On January 31, 1975, the Court granted Sea-crest’s motion for separate trials under Fed.R. Civ.P. 42(b). The Court ordered “that the issue of the validity of and the proprietary right, if any, of the plaintiff, Texas Urethane, Inc., in and to the formulation for polyurethane foam which it claims to be a trade secret shall be tried first and separate and apart from all other issues .

. Judge Noel had requested that all Galveston Division cases be assigned to newly appointed Judge Sterling, the presiding judge for that Division.

. Seacrest has gone out of business; its attorneys waived oral argument.

. Witnesses alluded to various reasons for the sudden change. Seacrest was plagued by many scrap rejects (which were discarded), production rejects (which were sold at a lower price or burned), and warranty claims (which were returned). Moreover, there was a significant variance in the composition of batches sent to Seacrest by Urethane. This made production of boats difficult. There was also some concern about quality control and the amount of production facilities at Urethane’s plant.

. On February 20, 1974, a Stipulation of Dismissal of Urethane’s claims against Carbide was entered.