Design Defect Claims Under Missouri Law

15 Min Read

The undefined standard, the cases that built it, and how design cases are actually won and lost

A design-defect claim is the most conceptually ambitious species of products litigation. It does not accept the product’s blueprint and quarrel with a stray unit; it indicts the blueprint itself, contending that every unit the manufacturer built to specification is unreasonably dangerous. Missouri’s treatment of that claim is, in a word, idiosyncratic. The State has adopted strict liability in tort, has codified it, and has then—deliberately and repeatedly—refused to tell juries what the operative phrase means. The result is a body of law in which the outcome-determinative question, whether a product is “unreasonably dangerous,” is handed to the jury as a bare ultimate issue, and in which the real contest is fought not over the legal standard but over the evidence and argument that fill the vacuum the standard leaves. This article traces how that framework was built, states the elements and the instructions, canvasses the case law a Missouri design lawyer must know, and turns to how these cases are actually tried.

I. The Statutory Frame and the Two-Branch Structure

The starting point is section 537.760 RSMo, enacted as part of the 1987 tort-reform package (L. 1987 H.B. 700 § 33, effective July 1, 1987). It defines a “products liability claim” as one seeking damages on a strict-liability theory because:

(1) The defendant, wherever situated in the chain of commerce, transferred a product in the course of his business; and (2) the product was used in a manner reasonably anticipated; and (3) either or both of the following: (a) the product was then in a defective condition unreasonably dangerous when put to a reasonably anticipated use, and the plaintiff was damaged as a direct result of such defective condition as existed when the product was sold; or (b) the product was then unreasonably dangerous when put to a reasonably anticipated use without knowledge of its characteristics, and the plaintiff was damaged as a direct result of the product being sold without an adequate warning.

Two features of this text govern everything that follows. First, subsection (3) splits the tort into two independent branches—the defect branch of (3)(a), which houses both design and manufacturing defects under the single “defective condition unreasonably dangerous” formula, and the failure-to-warn branch of (3)(b). The Supreme Court has held those branches are genuinely distinct theories that may be submitted from the same product, a point developed below in the discussion of Moore v. Ford. Second, and easy to miss, “reasonably anticipated use” appears twice—as the free-standing element in subsection (2) and again inside the defect definition in (3)(a). That double appearance is not redundancy; it is doctrinally load-bearing, and it is why misuse analysis in Missouri is more nuanced than the label suggests.

Critically, section 537.760 codified the pre-existing common-law action; it did not rewrite it. The statute supplies the vocabulary of the § 402A “defective condition unreasonably dangerous” standard, and the Supreme Court has read it as a codification of that standard rather than as an invitation to import the risk-utility and reasonable-alternative-design regime of the Restatement (Third). A Missouri design lawyer therefore works from three sources at once: the Restatement (Second) § 402A, the Keener–Blevins–Nesselrode line of decisions, and Chapter 537.

II. From Keener to Blevins: How Missouri Acquired Strict Design Liability

Missouri entered the strict-liability era in Keener v. Dayton Electric Manufacturing Co., 445 S.W.2d 362 (Mo. 1969), a wrongful-death case in which a man was electrocuted lifting an ungrounded electric sump pump—sold without a ground wire or overload protector—while standing in water. The Court “adopt[ed] the rule of strict liability in tort stated in 2 Restatement, Law of Torts, Second, § 402A,” 445 S.W.2d at 364, reproducing the section in full, including its two hallmark features: liability attaches although “the seller has exercised all possible care,” and it requires that the product reach the user “without substantial change in the condition in which it is sold.” Keener, it bears noting, was a manufacturing-process case; the Court said so itself years later in Nesselrode. The design branch had to be added.

It was added in Blevins v. Cushman Motors, 551 S.W.2d 602 (Mo. banc 1977), a three-wheeled golf-cart rollover that threw and crushed the plaintiff. The Court held that § 402A reaches defective design and not merely manufacturing flaws—“there is no rational distinction between design and manufacture,” 551 S.W.2d at 607—and located the inquiry on “the condition or character of the product and not the character of the defendant’s conduct,” id. at 607–08. Blevins drew its analytical rationale from Oregon’s Phillips v. Kimwood Machine Co., framing the question as whether a reasonably prudent manufacturer with knowledge of the product’s dangerous condition would have placed it on the market, and it treated evidence of feasible safer alternatives—widening the wheelbase, adding a fourth wheel, lowering the center of gravity, reducing speed—as sufficient to submit the case. A caution for the practitioner: Blevins is sometimes misdescribed as adopting an enumerated Wade-factor risk-utility test. It does not. It embraces risk-utility reasoning as a mode of analysis without codifying a fixed factor list, and that distinction is exactly what Nesselrode would later make explicit.

III. Nesselrode and the Deliberate Non-Definition

Nesselrode v. Executive Beechcraft, Inc., 707 S.W.2d 371 (Mo. banc 1986), is the keystone of Missouri design law, and it repays close reading. A Beech Baron crashed on climb-out near Kansas City, killing all aboard. During a routine inspection weeks earlier, worn elevator trim-tab actuators had been replaced; the left and right actuators were visually identical but functionally distinct, and the mechanics installed them reversed, so the trim tabs moved opposite the pilot’s inputs. The plaintiffs’ design theory was that the actuators lacked “murphy-proof” asymmetric keying—physical features permitting only correct installation—despite Beech’s own engineering manual calling for such features “in applications where the consequence of wrong assembly presents any hazardous condition.” The plaintiffs abandoned negligence and submitted solely on strict-liability design defect and failure to warn. The jury returned $1.5 million; the Court of Appeals ordered a directed verdict for Beech; the Supreme Court reversed the Court of Appeals and reinstated the verdict.

In doing so, the Court delivered the passage that defines the field:

Under our model of strict tort liability the concept of unreasonable danger, which is determinative of whether a product is defective in a design case, is presented to the jury as an ultimate issue without further definition.

And it explained, with unusual candor, that it was choosing neither of the competing tests:

Though Missouri has adopted the rule of strict tort liability as set forth in the Restatement, we have not yet formally incorporated, in any meaningful way, the Restatement’s consumer expectation test into the lexicon of our products liability law. . . . Nor have we yet decided to travel or require plaintiffs to travel the path of risks and utilities. . . . Because of these two decisional constraints, we choose not to decide this question and instead, limit our discussion to edificatory ends.

The Court defended leaving the term bare on a jury-competence rationale drawn from Professor Leon Green: juries resolve comparably open normative concepts—“negligence,” “reasonableness”—without fictitious standards, and importing “abstract, abstruse standards” risks perpetuating “the mystical trial by ordeal.” Nesselrode at 707 S.W.2d 371, 377–78. The opinion also settled two subsidiary points that recur constantly. It stated the design inquiry as whether the product, “because of the way it was designed, creates an unreasonable risk of danger to the consumer or user when put to normal use,” with the plaintiff bearing the burden to prove the product “unreasonably dangerous and therefore ‘defective.’” And it held that “reasonably anticipated use . . . includes misuse and abnormal use which is objectively foreseeable,” id. at 380–81, so that the mechanics’ negligent reversed installation was a foreseeable misuse rather than a superseding cause. The takeaway: in Missouri there is no test the plaintiff must satisfy as a matter of law and none the defendant may demand by instruction. The concept is the jury’s to give content.

IV. The Modern Supreme Court: Rodriguez II and Moore

Two later Supreme Court decisions confirm and sharpen the Nesselrode frame. The first is Rodriguez v. Suzuki Motor Corp., 996 S.W.2d 47 (Mo. banc 1999)—decided by the Court en banc, not, as it is sometimes miscited, by the Court of Appeals. It arose from a Suzuki Samurai rollover and a strict-liability crashworthiness theory. The Court reversed a plaintiff’s judgment on evidentiary grounds—holding that authenticated government safety reports were admissible under section 490.220 RSMo and that the defense was entitled to wide latitude to cross-examine the plaintiff’s rollover-simulation expert about a methodology internally derided as “practically worthless”—but for present purposes its doctrinal holding is what matters: the Court declined to judicially redefine “unreasonably dangerous” or to adopt the Restatement (Third)’s risk-utility and reasonable-alternative-design test, holding that section 537.760 codified the § 402A standard and that any change was for the legislature. Rodriguez is the modern anchor for Missouri’s refusal to follow the Third Restatement.

The second is Moore v. Ford Motor Co., 332 S.W.3d 749 (Mo. banc 2011), which supplies the cleanest statements of two principles. A driver’s seatback collapsed rearward in a rear-end collision, rendering her paraplegic; she pleaded both design defect and failure to warn. On the reasonable-alternative-design question, the Court held flatly: “Missouri does not require a plaintiff to create an alternative design to prove a design defect claim; it is enough that plaintiff show that the design used was defective and unreasonably dangerous.” Id. at 760. On the relationship between the theories, the Court held that design defect and failure to warn “constitute distinct theories aimed at protecting consumers from dangers that arise in different ways,” and that a jury could find “a product’s design is not unreasonably dangerous in itself but that, without an accompanying warning . . . the otherwise non-defective product is unreasonably dangerous.” Id. at 757. That holding—developed in the failure-to-warn article in this series—means a design defense verdict does not moot a warning claim, and it invites double submission from a single product.

V. The Eighth Circuit’s Consumer-Expectations Gloss

Here is a tension every Missouri practitioner must hold in mind. Nesselrode formally adopts neither the consumer-expectation nor the risk-utility test—yet the federal courts applying Missouri law repeatedly resolve design cases as a matter of law using the § 402A comment i consumer-expectations formulation and open-and-obvious reasoning. There is, in other words, a gap between the formal “no test” rule and the functional consumer-expectations gloss that decides cases in practice. The leading illustration is Linegar v. Armour of America, Inc., 909 F.2d 1150 (8th Cir. 1990). A Missouri state trooper was killed when a bullet entered the unprotected side gap of a “contour” ballistic vest that had stopped five other rounds; the jury awarded $1.5 million, and the Eighth Circuit reversed and directed judgment for the manufacturer. The vest, the court held, was “neither defective nor unreasonably dangerous” as a matter of law: it “performed precisely as expected,” its coverage limits were “obvious . . . to anyone with ordinary knowledge, most especially the vest’s wearer,” and—words worth committing to memory—“[a] manufacturer is not obliged to market only one version of a product, that being the very safest design possible.” 909 F.2d at 1154–56.

The same register runs through Pree v. Brunswick Corp., 983 F.2d 863 (8th Cir. 1993) (intoxicated plaintiff injured by an unguarded boat propeller; defense verdict affirmed; strict liability “is not, nor was it ever intended to be, an enveloping net of absolute liability”), and Bachtel v. TASER International, Inc., 747 F.3d 965 (8th Cir. 2014) (design and warning claims failed on summary judgment after the plaintiff’s experts were excluded; liability turns on “the condition or character of the product and not the character of the defendant’s conduct”). The practical lesson is that although a Missouri plaintiff need not disprove consumer expectations or prove risk-utility, a design whose limitation is open, obvious, and disclosed—and that performs as an ordinary user would expect—remains highly vulnerable to judgment as a matter of law. The formal doctrine is plaintiff-friendly; the functional doctrine, in federal court especially, is not.

VI. Reasonably Anticipated Use, Foreseeable Misuse, and the Comparative-Fault Line

Because “reasonably anticipated use” is an affirmative element—appearing twice in the statute and twice in the verdict director—misuse operates on two distinct planes, and keeping them separate is central to both sides’ strategy. Foreseeable misuse does not defeat the claim; it is subsumed within reasonably anticipated use. Jarrell v. Fort Worth Steel & Manufacturing Co., 666 S.W.2d 828 (Mo. App. E.D. 1984), is the workhorse authority: a maintenance chief lost his arm in an ice conveyor whose access-door placement forced workers to reach into the auger housing, and the court held the manufacturer “should have reasonably anticipated that maintenance people would insert their arms,” so that “[w]here there is both a design defect and misuse . . . the misuse does not become an intervening cause if the misuse was foreseeable.” The design duty thus extends to foreseeable servicing and maintenance conditions.

The line that matters is this: misuse so far outside anticipation that it negates the reasonably-anticipated-use element is a complete defense (the plaintiff fails to make a submissible case), whereas misuse that is foreseeable but careless is partial—it is routed to pure comparative fault under section 537.765 RSMo and reduces damages without barring recovery. The doctrinal weight of the element is illustrated by Bayes v. Biomet, Inc., 55 F.4th 643 (8th Cir. 2022), where a jury rejected strict-liability defect but found negligent design in a metal-on-metal hip-implant case. The Eighth Circuit held the split verdict reconcilable precisely because strict liability required proof of reasonably anticipated use while negligence did not—a reminder that the element is not a formality but the hinge on which mixed verdicts turn.

VII. Alternative Design: Powerful Evidence, Not an Element

It follows from the undefined standard that Missouri does not make a feasible safer alternative design an element of the claim; Moore says so at 332 S.W.3d at 760, and the Eighth Circuit, applying Missouri law, has confirmed that a plaintiff need not prove a reasonable alternative design to reach the jury. This places Missouri outside the group of jurisdictions—Texas prominent among them—that have adopted the Restatement (Third) § 2(b) reasonable-alternative-design requirement. But “not an element” is a long way from “unimportant.” In Nesselrode itself the feasibility of the murphy-proof alternative was the backbone of the plaintiffs’ proof, and it nearly always is, for reasons of persuasion, causation, and appellate durability that are treated at length in the companion article, Safer Alternative Design Requirements. The point to carry into a design case is that the plaintiff who declines to prove an alternative is exercising a legal right and, usually, committing a tactical error.

VIII. The Verdict Director: MAI 25.04 and the Missing Definition

The strict-liability design claim is submitted under MAI 25.04, whose four conjunctive findings track the statute: that the defendant sold the product in the course of its business; that the product was then in a defective condition unreasonably dangerous when put to a reasonably anticipated use; that it was used in a manner reasonably anticipated; and that the plaintiff was damaged as a direct result of the defective condition as it existed when the product was sold. The single most important structural fact about the instruction is what it omits: there is no definitional paragraph for “unreasonably dangerous”—no bracketed definition, no risk-utility factor list, no consumer-expectation gloss. That absence is the instructional embodiment of Nesselrode, and it is why the defense cannot obtain a defining instruction. In Smith v. Brown & Williamson Tobacco Corp., 275 S.W.3d 748 (Mo. App. W.D. 2008), the defendant tendered an instruction defining “unreasonably dangerous”; its refusal was not error, because the term is deliberately left to the jury. The same MAI 25.04 template serves both design and manufacturing defect—the elements are identical, and the distinction lives in the evidence, not the instruction.

Two companion instructions round out the picture. Negligent design is submitted on a negligence verdict director (generally MAI 25.09), which focuses on the reasonableness of the manufacturer’s conduct rather than the condition of the product—the fork that Bayes and Bavlsik v. General Motors, LLC, 870 F.3d 800 (8th Cir. 2017) (recognizing negligent failure to test a design as a viable Missouri theory), show can yield a plaintiff’s verdict even where strict liability fails. And the comparative-fault converse, MAI 32.23, keyed to the section 537.765(3) fault categories, must hypothesize specific facts rather than parrot the statutory language, on pain of submitting a “roving commission.” Because MAI is copyrighted and periodically renumbered, confirm the current instruction numbers, verbatim text, and Notes on Use against the official MAI volume before drafting.

IX. Experts and the 2017 Daubert Regime

Design cases are expert cases, and since 2017 the expert has had to clear a higher bar. Effective August 28, 2017 (L. 2017 H.B. 153), section 490.065.2 RSMo replaced Missouri’s prior, more permissive standard with language identical to Federal Rule of Evidence 702: the opinion must help the trier of fact, rest on sufficient facts or data, be the product of reliable principles and methods, and reflect a reliable application of those methods to the facts. Because the text mirrors Rule 702, Missouri courts treat federal Daubert, Kumho Tire, and Joiner authority as persuasive, and the gatekeeping bite in products cases is real—see Gebhardt v. American Honda Motor Co., 627 S.W.3d 37 (Mo. App. W.D. 2021), affirming exclusion of a sudden-acceleration causation expert whose opinions rested on untested assumptions with “too great an analytical gap between the data and the opinion proffered.”

Two practice points deserve emphasis. First, preservation is dispositive: Peters v. General Motors Corp., 200 S.W.3d 1 (Mo. App. W.D. 2006), holds that a section 490.065 reliability challenge is waived unless renewed by trial objection—a motion-in-limine ruling is interlocutory—and that once an opinion is admitted it supports submissibility. Second, exclusion can be case-dispositive; Bachtel affirmed summary judgment after the plaintiff’s experts were struck. The doctrinal irony is worth naming: Missouri screens the reliability of “unreasonably dangerous” and alternative-design opinions rigorously even though the substantive standard those opinions address remains judicially undefined. Post-2017, the evidentiary gate increasingly does the disciplining work that the missing legal definition does not.

X. Trying the Design Case

Submissibility. The plaintiff receives the most favorable view of the evidence and all favorable inferences, and circumstantial evidence will support submission—but only if it “reasonably point[s] to the desired conclusion and tend[s] to exclude any other reasonable conclusion.” Peters, 200 S.W.3d at 8–9 (quoting Weatherford); evidence pointing “equally” to a non-actionable cause is insufficient, Kircher v. Purina Mills, 775 S.W.2d 115, 117 (Mo. banc 1989). Directed-verdict and JNOV wins for the defense tend to come on causation and on the “beyond the ordinary consumer’s contemplation”/obviousness points illustrated by Linegar and Pree.

Exploiting the undefined standard. For the plaintiff, the absence of any definition is an asset: it permits argument of “unreasonably dangerous” in ordinary, intuitive, consumer-safety terms, unconstrained by a risk-utility burden or a mandatory alternative-design element, and it justifies resisting any defense-tendered gloss (Smith). For the defense, the same vacuum must be filled on the evidence-and-argument plane rather than by instruction: import consumer-expectation and open-and-obvious themes (Linegar), press state-of-the-art and feasibility proof, marshal government and industry standards (authenticated federal reports are admissible under section 490.220, per Rodriguez), and attack submissibility, causation, and expert reliability under section 490.065. The recurring instruction fight—whether any gloss on “unreasonably dangerous” can be injected—Nesselrode and Smith answer in the negative, which structurally favors plaintiffs at the instruction stage and pushes the decisive contest into the evidence and the closing.

Defense themes to preserve. Beyond obviousness and “not the safest possible design” (Linegar), the standard defense package includes alteration or substantial change (the defect must exist “as existed when the product was sold”), misuse pressed first as a complete defeat of reasonably anticipated use and, failing that, as section 537.765 comparative fault, state-of-the-art evidence bearing on dangerousness, and, in device and drug cases, the learned intermediary and sophisticated-user doctrines. Each should be pleaded and, where supported, submitted with fact-specific hypotheses rather than statutory abstractions.

XI. Analytical Tensions and Open Questions

Three tensions are worth watching. The first is the instability inherent in an undefined ultimate issue: Missouri asks lay juries to apply “unreasonably dangerous” with no legal content, trading doctrinal predictability for jury latitude and making appellate review of submissibility comparatively opaque. The second is the gap, canvassed above, between the formal “no test” rule and the functional consumer-expectations gloss the Eighth Circuit deploys to decide cases as a matter of law—a divergence that can make forum selection outcome-affecting. The third is the awkward pairing of a rigorous post-2017 Daubert regime with a substantive standard that has no definition; expect defense arguments that a plaintiff’s “unreasonably dangerous” opinion is unreliable precisely because it rests on no defined methodology, and watch whether any court converts risk-utility from a permitted mode of evidence into a required element. To date the controlling authority has resisted that move—Rodriguez and Moore hold the line, and Bayes decided a modern device case without risk-utility balancing—but the pressure is real, and it is briefable. For now, the Missouri rule is stable and distinctive: the jury decides what “unreasonably dangerous” means, and the lawyer who best fills that vacuum with evidence and argument, on either side, is the one who wins.

Related in the Products Liability Series

Tags: