Discovery Scope and the Work Product Doctrine Explained

Under Rule 26(b)(1), a party may obtain discovery of any nonprivileged matter that is relevant to any party's claim or defense and proportional to the needs of the case. Those are two separate requirements joined by "and" — a request that is relevant but disproportionate is outside the scope, and so is a proportionate request for something irrelevant.

Relevance here is measured against the pleadings, not against some general sense of what might be interesting. The claims and defenses actually pleaded define the universe. That is why a motion to dismiss and a discovery fight are connected: narrowing the claims narrows what the other side can demand.

The rule then adds a line students should quote verbatim: information within this scope need not be admissible in evidence to be discoverable. Discoverability is broader than admissibility. But the boundary of the broader circle is relevance plus proportionality — admissibility is simply not the test in either direction.

Rule 26(b)(2)(C) backs this up with a mandatory limit. The court must limit discovery when it is unreasonably cumulative or duplicative, when it can be obtained from a more convenient or less burdensome source, when the requesting party has already had ample opportunity to get it, or when it falls outside the Rule 26(b)(1) scope. A court can impose these limits on its own, not only when a party objects.

Proportionality is not a vibe; Rule 26(b)(1) lists six specific factors, and a good answer walks through them. They are: the importance of the issues at stake in the action; the amount in controversy; the parties' relative access to relevant information; the parties' resources; the importance of the discovery in resolving the issues; and whether the burden or expense of the proposed discovery outweighs its likely benefit.

Two of these are frequently misread. "Importance of the issues at stake" is not a synonym for the dollar figure — it is a separate factor precisely because some cases (civil rights, injunctive relief, matters of public significance) matter beyond their monetary value. And "the parties' relative access to relevant information" targets information asymmetry: when one side holds nearly all the documents, that imbalance weighs in favor of allowing the discovery, not against it.

No single factor controls, and the burden is shared. Proportionality is not a device that lets the responding party simply announce that a request is too expensive; nor does it let the requesting party ignore burden because its claim is large. Courts expect both sides to address the factors with something more concrete than adjectives.

The 2015 amendments to Rule 26 did two things worth knowing cold. First, they moved the proportionality factors out of Rule 26(b)(2)(C)(iii), where they had operated as a limit a court could apply, and into the definition of scope in Rule 26(b)(1). Proportionality is now part of what discovery is, not an exception to it.

Second — and this is the point 1Ls keep getting wrong — the amendments deleted the phrase "reasonably calculated to lead to the discovery of admissible evidence." That clause had never been intended as a definition of scope. It existed to make clear that inadmissibility alone is not a ground to object. Courts and lawyers nonetheless quoted it as if it were an independent, broader test for what could be demanded. The rule now says plainly that information need not be admissible to be discoverable, and the old phrase is gone.

The amendments also removed the provision that had allowed a court, on a showing of good cause, to order discovery of matter merely "relevant to the subject matter involved in the action." The single operative standard today is relevance to a claim or defense, plus proportionality. If your exam answer still recites "reasonably calculated," you are quoting a repealed standard.

Work product protects materials prepared for litigation from discovery, even when they are plainly relevant and proportional. The doctrine comes from Hickman v. Taylor (1947), where the Supreme Court refused to let a party obtain an opposing lawyer's witness interview notes and memoranda absent a showing of necessity, reasoning that an attorney must be able to prepare a case without the adversary rifling through the file. The doctrine was later codified for documents and tangible things in Rule 26(b)(3).

The rule has three elements. The protection covers (1) documents and tangible things, (2) prepared in anticipation of litigation or for trial, (3) by or for another party or that party's representative — and the rule expressly lists the attorney, consultant, surety, indemnitor, insurer, or agent. No lawyer is required. An insurance adjuster's litigation file can be work product.

The protection is qualified, not absolute. Under Rule 26(b)(3)(A)(ii), a party can still get the materials by showing substantial need for them to prepare its case and that it cannot, without undue hardship, obtain their substantial equivalent by other means. Both halves are required: need alone is not enough if the same information can be gathered another way.

One structural point people miss: work product protects materials, not facts. The other side can depose your witness and ask what she saw. What they cannot do is take the memorandum in which your lawyer wrote down what she said. Rule 26(b)(3)(C) adds a carve-out — a person may obtain their own previous statement about the action without any showing at all.

Most real work product disputes are won or lost on this element, not on substantial need. A document created in the ordinary course of business is not work product no matter who ends up holding it or how helpful it would be to the other side. Litigation need not have been filed, but it must have been reasonably anticipated — more than a remote or speculative possibility.

Courts describe the test in two ways. The formulation followed in most federal circuits asks whether the document was created "because of" the prospect of litigation — would it have been created in essentially similar form regardless? Some courts instead ask whether litigation was the "primary purpose" of the document. The formulations usually converge, but they can diverge on dual-purpose documents, and jurisdictions genuinely differ, so identify the standard rather than assuming one.

Dual-purpose documents are the battleground: the accident report that serves both a safety-review function and a litigation-defense function, the internal investigation that satisfies a regulator and prepares a defense. The practical tell courts look for is whether the party's own routine practice would have produced the document anyway.

The rule protects two tiers of material at two different strengths, and conflating them is a reliable way to lose points. Ordinary work product — sometimes called fact work product — is the raw material: witness statements, photographs, investigative reports, compilations of data gathered for the case. It gets the qualified protection described above and yields to a showing of substantial need plus undue hardship.

The classic substantial-need showing involves a statement or observation that cannot be recreated: a witness who has since died, disappeared, or become genuinely unavailable; a contemporaneous account taken at the scene before conditions changed; a photograph of a condition that has since been repaired. Mere convenience or the expense of doing one's own investigation does not qualify.

Opinion work product is different in kind. Rule 26(b)(3)(B) provides that even when a court orders discovery of fact work product, it must protect against disclosure of the mental impressions, conclusions, opinions, or legal theories of a party's attorney or other representative concerning the litigation. That is a mandatory instruction to the court, which is why opinion work product is described as receiving near-absolute protection. Courts differ on whether it can ever be reached, and most that entertain the possibility require an extraordinary showing well beyond substantial need.

The practical consequence is redaction. When a court orders production of an investigator's report, the factual observations may come out while the attached counsel commentary about which witness will hold up on cross stays covered. Treat the two tiers as separate questions with separate standards.

These are different doctrines with different origins, different holders, different scope, and different waiver rules. They overlap constantly in practice — a lawyer's litigation memorandum to a client can be both — which is exactly why students blur them. Keep the two analyses separate and run both.

The privilege protects confidential communications between attorney and client made for the purpose of seeking or providing legal advice. It exists to encourage candor with counsel, so it attaches the moment legal advice is sought and requires no litigation at all. Work product protects materials generated because of litigation, communication or not, and requires anticipated litigation as a threshold.

Note the waiver line especially. Because confidentiality is an element of the privilege, handing a privileged memo to an outside consultant can destroy it. Work product asks a different question — whether the disclosure undercut the purpose of keeping material from the adversary — so sharing with a non-adversary often preserves it. In federal proceedings, Fed. R. Evid. 502 governs inadvertent disclosure and the reach of subject-matter waiver for both doctrines.

A delivery van strikes a pedestrian. Within hours, the company's risk manager photographs the intersection and takes signed statements from the two employees who witnessed it. Two weeks later, outside counsel writes a memorandum assessing which witness will hold up on cross-examination and estimating the company's exposure. The pedestrian sues and requests all three items.

Start with the photographs and statements. The threshold question is not substantial need — it is whether they were prepared in anticipation of litigation. If the company documents every incident this way as a matter of routine, they are ordinary-course business records, not work product at all, and they come out if relevant and proportional. If the company acted because a serious injury made litigation reasonably anticipated, they are ordinary work product.

Assume they are work product. The plaintiff must show substantial need plus inability to obtain the substantial equivalent without undue hardship. Statements taken hours after the collision from a witness who has since become unavailable, or whose memory has demonstrably faded, are the paradigm case for substantial need — that account cannot be recreated. Photographs of an intersection that has since been resurfaced make the same argument. If both witnesses are available and the intersection is unchanged, the plaintiff can simply take depositions and go look, and the showing fails.

Counsel's memorandum is a different tier. Assessments of witness credibility and case exposure are mental impressions, conclusions, and legal theories — opinion work product. Even if the plaintiff wins on the witness statements, Rule 26(b)(3)(B) directs the court to protect the memorandum's opinion content, so it is withheld or produced in redacted form. And if the memorandum was a confidential communication to the company's decision-makers seeking legal advice, it is very likely privileged as well; the company would assert both grounds.

Finally, note what is not protected. The two employees' own knowledge of the accident is a fact, and facts are always fair game. The plaintiff can depose them and ask exactly what they saw, when, and from where. Rule 26(b)(5) then requires the company to serve a privilege log describing what it withheld in enough detail for the plaintiff to assess each claim without revealing the protected content itself.

First, writing "reasonably calculated to lead to the discovery of admissible evidence." That standard was deleted in 2015. The scope test is relevance to a claim or defense plus proportionality, and Rule 26(b)(1) separately states that information need not be admissible to be discoverable.

Second, treating relevance as the whole inquiry and mentioning proportionality as an afterthought. Since 2015 it sits inside the definition of scope. An answer that names the six factors and applies two or three of them to the facts will outscore one that says "this seems burdensome."

Third, using "privileged" as a loose synonym for "protected." Attorney-client privilege and work product are separate doctrines with separate elements. Run both, and say which one you are applying.

Fourth, forgetting that work product protects materials rather than facts. If an answer says the other side cannot learn what a witness observed because counsel wrote it down, that is backwards.

Fifth, skipping the ordinary-versus-opinion split. The standard changes: fact work product yields to substantial need plus undue hardship, while opinion work product must be protected even when the court orders the rest of the file produced. And a framing note — this is a study guide to how courts describe the doctrine, not advice about any actual discovery dispute. Real fights turn on the local rules, the case law of the particular circuit, and the specific documents in front of the judge.