JUSTICE DUBOFSKY delivered the opinion of the Court.

We granted certiorari to review the unpublished opinion of the Court of Appeals in Hansen v. Belle Bonfils Memorial Blood Bank, No. 80CA1173 (September 10, 1981), which held that the exception to strict liability for unavoidably unsafe products does not apply to transfused blood contaminated by hepatitis virus. We reverse the judgment of the Court of Appeals and remand the case for further proceedings.

The plaintiff-respondent, Muriel Hansen, contracted hepatitis in early 1971 after receiving a transfusion of blood supplied by the defendant-petitioner, Belle Bonfils Memorial Blood Bank (the Blood Bank). Hansen and her husband brought this action for negligence, breach of implied warranties, and strict liability in tort against Mercy Hospital and the Blood Bank, claiming that the transfused blood contained the virus which caused her hepatitis. Relying on St. Luke's Hospital v. Schmaltz, 188 Colo. 353, 534 P.2d 781 (1975), the district court dismissed the claims, finding that the supplying of blood was a service, not a sale of a product. This ruling was reversed as to claims against the Blood Bank in Hansen v. Mercy Hospital, 40 Colo. App. 17, 570 P.2d 1309 (1977), aff'd sub nom. Belle Bonfils Memorial Blood Bank v. Hansen, 195 Colo. 529, 579 P.2d 1158 (1978), in which the Court of Appeals held that blood banks, unlike hospitals, are primarily in the business of selling blood, and thus remain amenable to suit for breach of warranty and strict liability in tort. Accord, Russell v. Community Blood Bank, 185 So.2d 749 (Fla.Dist.Ct.App. 1966), aff'd as modified 196 So.2d 115 (Fla. 19 7); Brody v. Overlook Hospital, 127 N.J.Super. 331, 317 A.2d 392, aff'd 66 N.J. 448, 332 A.2d 596 (1975); Carter v. Inter-Faith Hospital, 60 Misc.2d 733, 304 N.Y.S.2d 97 (Sup.Ct. 1969).

On remand, the Blood Bank moved for summary judgment on the warranty and strict liability claims on the ground that contaminated blood falls within an exception to the rule making a seller strictly liable for damage caused by a defective product. Under this exception, known as comment k to § 402A of the Restatement of Torts, the manufacturer or seller of a product which is vitally important yet unavoidably unsafe is not held strictly liable when it can prove that the product's preparation, marketing, and accompanying warnings were carried out in conformance with the highest known scientific and technical standards. See Restatement (Second) of Torts (1965) § 402A, comment k.

The Blood Bank submitted the uncontroverted affidavits of two doctors who specialized in the area of blood banking. The affidavits established that in 1971 there was an unavoidable risk of hepatitis associated with blood transfusions, a risk commonly known in the medical profession and incapable of being eliminated. According to the affidavits, there are two categories of viruses which can cause serum hepatitis: hepatitis B and hepatitis non-A non-B. In 1971, there was no way to detect hepatitis non-A non-B, the strain with which Hansen probably was infected. Hepatitis B was discoverable by testing blood samples prior to transfusion only 26% of the time. Therefore, the affiants stated, the only established methods for reducing the risk of hepatitis lay in the screening of donors, monitoring the results of post-transfusion hepatitis, and tracing contaminated blood to particular donors where possible.

The district court held that transfused blood containing hepatitis virus was an unavoidably unsafe product which merited exemption from strict liability under comment k, that the unavoidably unsafe product designation constituted a defense to the claim for breach of implied warranties, and that evidence that the contaminated blood had been processed according to state of the art techniques was a separate, absolute defense to both products liability claims. The Court of Appeals reversed the comment k and the absolute defense rulings. First, adopting a strict construction of comment k, it held that the exception did not apply.

"Comment k acknowledges only that a product may present a justifiable risk of harm even in the absence of a defect or some adulterating substance. Such products are neither defective nor unreasonably dangerous. However, blood supplied for the purpose of transfusion and which contains hepatitis virus is a defective product, dangerous to all. Hence, the comment k exception does not apply here."

Hansen v. Belle Bonfils Memorial Blood Bank, No. 80CA1173 (September 10, 1981), slip op. at 2. The dissent argued that comment k should have been applied here. Second, the Court of Appeals held that evidence of state of the art processing of blood could not operate as a separate defense to either claim. Based on its assumption that the blood at issue was defective and unreasonably dangerous, it found that state of the art evidence would be irrelevant.

"To hold otherwise would inject a standard of care analysis into the case, but the policy underlying 402A strict liability is that the seller of a product is liable even though he has exercised all possible care in the preparation and sale of his product. Hence, due care is not an issue."

Id., slip op. at 3.

In our view, the Court of Appeals erred in failing to apply the comment k exception to the rule of strict liability in tort for defective products, but correctly recognized that state of the art evidence does not constitute a separate, absolute defense to strict liability for a product such as contaminated blood. In Part I of this opinion, we discuss the comment k exception to strict liability in tort; in Part II, the relevance of state of the art evidence to the comment k defense; in Part III, application of comment k to the claim for breach of implied warranties; and in Part IV, the application of these principles to this case.

I

The elements of cause of action for strict liability are set out in Restatement (Second) of Torts (1965) § 402A. This Court adopted § 402A in Hiigel v. General Motors Corp., 190 Colo. 57, 544 P.2d 983 (1975). Comment k to that section precludes liability for unavoidably unsafe products:

"There are some products which, in the present state of human knowledge, are quite incapable of being made safe for their intended and ordinary use. These are especially common in the field of drugs. An outstanding example is the vaccine for the Pasteur treatment of rabies, which not uncommonly leads to very serious and damaging consequences when it is injected. Since the disease itself invariably leads to a dreadful death, both the marketing and the use of the vaccine are fully justified, notwithstanding the unavoidable high degree of risk which they involve. Such a product, properly prepared, and accompanied by proper directions and warning, is not defective, nor is it unreasonably dangerous. The same is true of many other drugs, vaccines, and the like, many of which for this very reason cannot legally be sold except to physicians, or under the prescription of a physician. It is also true in particular of many new or experimental drugs as to which, because of lack of time and opportunity for sufficient medical experience, there can be no assurance of safety, or perhaps even of purity of ingredients, but such experience as there is justifies the marketing and use of the drug notwithstanding a medically recognizable risk. The seller of such products, again with the qualification that they are properly prepared and marketed, and proper warning is given, where the situation calls for it, is not to be held to strict liability for unfortunate consequences attending their use, merely because he has undertaken to supply the public with an apparently useful and desirable product, attended with a known but apparently reasonable risk." (Emphasis in original.)

As the illustrations in comment k indicate, the exception was intended to apply to drugs and medical products. Kimble Lesher, Products Liability § 376 (1979); 3 Frumer Friedman, Products Liability § 33.02 [4] (1983); Robb, A Practical Approach to Use of State of the Art Evidence in Strict Products Liability Cases, 77 Northwestern L.Rev. 1, 16 (1982).

The applicability of comment k to a given product depends upon the co-existence of several factors which justify the product's exception from strict liability. The product's utility must greatly outweigh the risk created by its use; the risk must be a known one; the product's benefits must not be achievable in another manner; and the risk must be unavoidable under the present state of knowledge. Phillips, A Synopsis of the Developing Law of Products Liability, 28 Drake L.Rev. 317, 357 (1978); Murray, The State of the Art Defense in Strict Products Liability, 57 Marquette L.Rev. 649, 656 (1974). It is the manufacturer's burden to establish all of the conditions for the application of comment k. Id. at 657.

Thus, in order to fall within the comment k exception, not only should an unavoidably unsafe product carry a unique or profound benefit, but that benefit should extend to the vast majority of the users of the product. "It does not serve society that an unavoidably unsafe product, which has occasional or fractious benefit, should enjoy insulation from strict liability in tort when the product's predominant effects are detrimental to individual and public safety." Willig, The Comment K Character: A Conceptual Barrier to Strict Liability, 29 Mercer L.Rev. 545, 545 (1978).

A further condition for the applicability of comment k is that the risk be known to the manufacturer at the time of distribution: "The seller of such products . . . is not to be held to strict liability . . . merely because he has undertaken to supply the public with an apparently useful and desirable product, attended with a known but apparently reasonable risk." Restatement (Second) of Torts (1965) § 402A, comment k (emphasis added). Comment k does not apply to unknown risks associated with products defended as "unavoidably unsafe." Phillips, supra at 357. Knowledge of a product's dangerous tendencies is imputed to the manufacturer. Freund v. Cellofilm Properties, 87 N.J. 229, 432 A.2d 925 (1981); Spradley, Defensive Use of State of the Art Evidence in Strict Products Liability, 67 Minnesota L.Rev. 343, 395-6 (1982); Wade, On the Nature of Strict Tort Liability for Products, 44 Mississippi L.J. 825 (1973). The known risks then must be the subject of a warning. M. Dixon, Drug Product Liability, § 9.01[2] (1982). In the case of blood which is contaminated by hepatitis virus, the risk which is known is that of the potential danger involved in a unit of blood, not the exact condition of any particular unit. Wade, supra at 844-5.

Finally, the manufacturer must demonstrate that at the time of the preparation or marketing of the product, the state of the art had not progressed to where the risk was no longer unavoidable, and that the product's benefits could not be achieved by a substitute product or in another manner. Murray, supra, at 657. See Part II of this opinion.

In rejecting the application of comment k to this case, the Court of Appeals relied on Cunningham v. MacNeal Memorial Hospital, 47 Ill.2d 443, 266 N.E.2d 897 (1970), which approved a strict liability claim against a hospital for supplying blood contaminated with hepatitis. There, the Illinois Supreme Court considered the hospital's comment k defense but refused to adopt it. Instead, Cunningham distinguished comment k as applying "only to products which are not impure" — a label which could not be attached to hepatitis-contaminated blood. 266 N.E.2d at 904. As noted in the Court of Appeals' dissent, the classification of blood as impure in Cunningham drew a line between it and the Pasteur vaccine for prevention of rabies, the prototypical unavoidably unsafe product. The plaintiffs in this case adhere to the Cunningham pure-impure rationale, taking the distinction one step further. They argue that the two products are different because hepatitis-contaminated blood is dangerous to all recipients, while Pasteur vaccine has an adverse effect only upon those few individuals who are sensitive to it. Hence, the plaintiffs conclude, a blood transfusion involves a defective product, while Pasteur vaccine involves a "defective" recipient.

The affidavits submitted in support of the Blood Bank's summary judgment motion state that the original Pasteur vaccine was cultured on the brain tissue of laboratory animals or duck embryos. A small amount of the culture tissue cannot be filtered out of the vaccine, creating the risk of encephalitis in those individuals who have an idiosyncratic reaction to treatment with the vaccine. Thus, like contaminated blood, Pasteur vaccine is impure, and its use may have adverse consequences which cannot be predicted or avoided.

An additional problem with the Cunningham analysis is that it drew support from cases upholding strict liability claims for canned clams, bottled drinks, and similar products, the packaging of which prevents detection of defects. 266 N.E.2d at 902-3. The analogy to packaged goods is unpersuasive for two reasons. First, when a patient needs a blood transfusion, there is no real choice on the part of a physician to order it, a hospital or blood bank to supply it, or a patient to accept it. To the contrary, the production and purchase of consumer goods are, for the most part, volitional. See 16 Villanova L.Rev. 983, 996-7 (1971). Second, the raison d'etre of strict liability is to force some hazardous products out of the market. The same rationale does not apply to blood or vaccines which are life-saving and which have no known substitutes. Id. at 997-999.

Cunningham has been criticized both for the artificiality of the pure-impure analysis and for an overly-narrow construction of comment k. With rare exceptions, Cunningham has not been followed. Fogo v. Cutter Laboratories, 68 Cal.App.3d 744, 137 Cal.Rptr. 417 (1977); Fisher v. Sibley Memorial Hospital, 403 A.2d 1130 (D.C. 1979); McMichael v. American Red Cross, 532 S.W.2d 7 (Ky. 1975); Moore v. Underwood Memorial Hospital, 147 N.J.Super. 252, 371 A.2d 105 (1977); Hines v. St. Joseph's Hospital, 86 N.M. 763, 527 P.2d 1075, cert denied 87 N.M. 111, 529 P.2d 1232 (1974); see Spradley, supra at 388 ("a distinction between these situations is highly artificial, and should not have dictated the result . . . ."); Note, Strict Liability for Disease Contracted from Blood Transfusion, 66 Northwestern Univ. L.Rev. 80 (1971); 69 Michigan L.Rev. 1172 (1971); 16 Villanova L.Rev. 983 (1971); 24 Vanderbilt L.Rev. 645 (1971). But see Rostocki v. Southwest Florida Blood Bank, 276 So.2d 475 (Fla. 1973); DeBattista v. Argonaut-Southwest Ins. Co., 403 So.2d 26 (La. 1981); Reilly v. King County Central Blood Bank, 6 Wn. App. 172, 492 P.2d 246 (1971); 46 New York Univ. L.Rev. 403 (1971) ( Cunningham promotes the strict liability policy bases of risk allocation and incentive to improve product safety). Although commentators warned that Cunningham would usher in a new era of strict liability in medicine, its result effectively has been overruled by legislation in the vast majority of states, including Illinois ( see footnote 2, supra).

II.

The Blood Bank argues that evidence it complied with the state of the art for processing blood constitutes an absolute defense to a claim premised upon strict liability. We agree with the Court of Appeals that it does not, but for different reasons.

Evidence that a product conformed to the state of the art has been used in a strict liability action under § 402A to show that the product was not defective. Roberts v. May, 41 Colo. App. 82, 583 P.2d 305 (1978); Bruce v. Martin-Marietta Corp., 544 F.2d 442 (10th Cir. 1976); Murray, supra at 653. There are many definitions of "state of the art." For example, "state of the art" has been defined as the point of scientific and technological advance with respect to a given product at the time of the product's manufacture and design. Wiska v. St. Stanislaus Social Club, Inc., 390 N.E.2d 1133, 1138 n. 8 (Mass.App.Ct. 1979); Robb, supra, at 5. See Colorado J.I.2d 14:24 ("`State of the art' means the best technical, mechanical, and scientific knowledge and methods which are practical and available for use in the design, manufacturing, testing, inspecting, packaging and labeling of products in the same or similar industry for the same or similar products, for the purpose of providing for the quality and safety of such products.") When a manufacturer asserts the applicability of comment k, it must adduce evidence that the product was produced, marketed, and accompanied by warnings that conformed to the state of the art. Id. See 2 Frumer Friedman, supra, § 16A[4][f][i].

The overarching purpose of comment k is to permit a defendant to show that the product is not " unreasonably dangerous." See Restatement (Second) of Torts (1965) § 402A, comment k. In those jurisdictions which have applied comment k in cases involving contaminated blood, the defendant tacitly conceded the product was dangerous and focused upon the reasonableness of that danger. E.g., Hines v. St. Joseph's Hospital, supra, and Fogo v. Cutter Laboratories, supra. The decisions emphasize that blood is vital, that it has no known substitute, and that the risk of contracting hepatitis as a result of a transfusion is relatively low.

It appears that, while the incidence of hepatitis associated with transfusions of whole blood is difficult to determine precisely, in 1971 medical authorities placed the figure at from .06 to 3 per cent. 69 Michigan L.Rev. 1172, 1174, n. 12 (1971); Note, Strict Liability for Disease Contracted from Blood Transfusion, 66 Northwestern L.Rev. 80, 91, n. 50 (1971). The fatality rate from hepatitis was from five to ten percent.

Id.; Franklin, Tort Liability for Hepatitis: An Analysis and a Proposal, 24 Stanford L.Rev. 439, 443, n. 27 (1972).

When transfusions of blood collected from paid rather than voluntary donors are considered separately, however, the incidence of hepatitis increases dramatically. The ratio of hepatitis cases resulting from transfusions of commercial blood to those resulting from transfusions of volunteer blood has been estimated to be at least ten to one. Id. at 444. Federal regulations now require that every container of donated blood be prominently labeled "paid donor" or "volunteer donor." 21 C.F.R. § 606.120 (b)(2) (1982). Several courts have indicated that the source of transfused blood is relevant to the plaintiff's prima facie case of negligence against a blood bank. Hoder v. Sayet, 196 So.2d 205 (Fla.Dist.Ct.App. 1967); Hutchins v. Blood Services of Montana, 161 Mont. 359, 506 P.2d 449 (1973); Gilmore v. St. Anthony Hospital, 598 P.2d 1200 (Okla. 1979). But see Underwood Memorial Hospital, supra. Similarly, evidence of the source of transfused blood and of the availability of other, less risky alternative sources bears upon the reasonableness of the risk and is properly part of a plaintiff's rebuttal to a comment k defense. Accord, Jackson v. Muhlenberg Hospital, 53 N.J. 138, 249 A.2d 65 (1969).

The Court of Appeals' holding that state of the art evidence was irrelevant stemmed from its conclusion, based on Cunningham, that hepatitis-contaminated blood was inherently defective. Therefore, it held that the defendant's introduction of evidence controverting the plaintiff's proof that the danger was unreasonable would only cloud the settled issue of strict liability by injecting negligence concepts. Because we determine that comment k may be a defense in a strict liability action, it follows that state of the art evidence offered in support of the defense is relevant and admissible. However, the Court of Appeals correctly noted the tension between concepts of negligence and strict liability in cases involving comment k.

The distinction between negligence and strict liability theories in cases involving drugs, vaccines, and blood is the difference between focusing on the reasonableness of the manufacturer's conduct in the case of negligence, usually measured by the standard of the ordinary prudent person, and on the product's condition in the case of strict liability. Werner v. Upjohn Co., 628 F.2d 848 (4th Cir. 1980) cert. denied 449 U.S. 1080, 101 S.Ct. 862, 66 L.Ed.2d 804 (1981); Russell v. Community Blood Bank, supra; Spradley, supra; see Hiigel v. General Motors Corp., supra, 544 P.2d at 922. When comment k is asserted as an affirmative defense in a strict liability action, consideration of the reasonableness of the danger inherent in the product brings the suit "to the threshold of a suit for negligence." Russell v. Community Blood Bank, supra at 755. The burden of proving the reasonableness of the danger — measured by the elements of comment k — shifts to the defendant. Id. Further, the defendant's proof must rise to a higher level than that of a plaintiff in a prima facie case of negligence. The manufacturer is held to the knowledge and experience of an expert in the field, McEwen v. Ortho Pharmaceutical Corp., 270 Ore. 375, 528 P.2d 522 (1974), and the product must conform to the highest standards of available scientific and technical knowledge — not merely that knowledge available to the particular manufacturer. 2 Frumer Friedman, supra, at 16A[4][f][i]; Robb, supra, at 1, 7, 16.

Thus, we conclude that state of the art evidence is relevant to the defendant's burden of proof on two elements of the comment k affirmative defense: that the product's preparation, marketing, and accompanying warnings conformed to the highest known scientific and technical standards, and that the product's benefits were not achievable in another manner.

III.

At argument on the Blood Bank's motion for summary judgment, the parties assumed that comment k would apply to the claim for breach of implied warranties if it applied to the claim for strict liability in tort under § 402A. The district court agreed without discussion of the point, as did the Court of Appeals. We believe that in this case the district court and Court of Appeals reached the correct result, but we do not concur in this undifferentiated application of comment k to implied warranty claims.

While a manufacturer's proof that a product is unavoidably unsafe as contemplated by comment k is relevant in some cases to rebutting proof that a product failed to meet implied warranties of merchantability or of fitness for a particular purpose, this is not always true. When, for example, implied warranty theories overlap with concepts of express warranty, the use of comment k would not be appropriate. For example, Uniform Commercial Code section 4-2-314(2)(f), C.R.S. 1973, provides that to be merchantable, goods must "[c]onform to the promises or affirmations of fact made on the container or label if any." Such promises or affirmations may also constitute express warranties, Phillips, supra, at 340, and to the extent that they do, a defendant may not use comment k to insulate itself from strict liability. See, e.g., Uniform Product Liability Act § 105(c), 44 Fed.Reg. 2996, 2998 (1979) (comment k does not apply even to unavoidably unsafe products when the seller expressly warrants that the product is free from danger). Similarly, a product might be unavoidably unsafe in its most common use or application but impliedly warranted as fit for some other particular purpose. Comment k would not preclude a finding that a manufacturer breached an implied warranty of fitness for such a particular purpose. See Uniform Commercial Code, section 4-2-315, C.R.S. 1973.

Under the facts of this case, the Blood Bank's establishment of the elements of comment k would operate as a defense to the claims for breach of implied warranties as well. Express warranties did not contravene implied warranties, nor was the blood used for any purpose other than its generally accepted one. There are no other circumstances here which would tend to distinguish the warranty theories from the claim under § 402A. Proof that the transfused blood was vital and without a substitute, yet unavoidably unsafe after state of the art preparation, is relevant to the implied warranty claim.

IV.

The Blood Bank's proof in this case established two of the elements of comment k: that the utility of a blood transfusion greatly outweighs the risks attending that procedure, and that the risk of hepatitis contamination in donated blood is a known one. To prevail on remand, the Blood Bank must adduce proof of the remaining elements of comment k. First, it must show that there was no product substitute for donated whole blood in 1971.

Second, the Blood Bank must prove that its testing of the blood, screening of blood donors, and warning to the administering physician conformed to the best methods known in 1971.

The judgment of the Court of Appeals is reversed and the case is remanded for further proceedings.