ICH Q7 Defines the API Starting Material — ICH Q11 Says That Definition Isn't the Whole Test

Ask a CMC team where GMP begins for their API, and most will point to ICH Q7's glossary: the starting material is whatever gets incorporated as a “significant structural fragment” into the finished molecule. That answer is correct and incomplete in the same sentence. ICH Q7's Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients does define the term that way — but ICH Q11's Questions and Answers document on selecting and justifying starting materials says plainly that the structural-fragment test, used alone, will probably not satisfy a reviewer. A company that stops at Q7's one-line definition has answered only one of the questions a starting material designation has to survive.

What ICH Q7 actually defines

ICH Q7's glossary sets the term precisely: an API starting material is a raw material, intermediate, or API used in the production of an API that is incorporated as a significant structural fragment into the structure of the finished API. Practically, that boundary decides where Q7's GMP expectations — documented process controls, quality oversight, batch records — begin to apply. Steps before it sit outside Q7's direct scope, though FDA has been clear that it still expects an appropriate level of control over the material's production leading into that point; the boundary excuses a company from Q7's specific requirements, not from having a controlled process. Critically, the boundary is not something a company simply asserts. It is proposed in the CTD Module 3 filing and lives or dies on the regulatory authority's review of that proposal.

Why the structural-fragment test isn't the whole test

ICH Q11's Q&A document is direct about the failure mode this article is named for: a proposed starting material justified solely because it is a significant structural fragment of the API probably will not be accepted. The structural-fragment test is necessary, not sufficient. Regulators weigh several other considerations alongside it before accepting a proposed boundary:

• How much process remains downstream. Enough synthetic steps need to sit after the proposed starting material for the applicant to actually demonstrate control over the API's critical quality attributes. A boundary drawn too close to the finished API leaves too little described process for a reviewer to assess.
• Where impurities form and how they're purged. Steps that meaningfully affect the API's impurity profile — including mutagenic impurity carryover under ICH M7 — generally need to sit inside the described process, not upstream of a boundary chosen to avoid disclosing them.
• Well-defined chemical properties. The proposed material should be a synthesized or isolated substance with a defined structure and testable acceptance criteria for identity and purity — not a mixture or a loosely characterized intermediate.
• Independent availability. A material that is commercially available, or that could reasonably be manufactured independently of the API process, supports the case that it is a genuine starting point rather than a convenient stopping point chosen to minimize disclosed process.

A structural fragment answers one question. Whether the process before it is controlled enough that regulators don't need to see inside it is a different question — and it's usually the one that gets a starting material designation rejected. Why ICH Q11's Q&A exists

What happens when the boundary doesn't hold

A starting material designation that fails on review does not fail quietly. A reviewer who is not satisfied pushes the boundary upstream, which typically means new process description, new validation data, and sometimes a filing amendment mid-cycle — expensive at any stage, and worse close to submission. The same exposure shows up during GMP inspections of small molecule and oral solid dosage programs: an inspector who does not accept the stated starting material effectively extends GMP expectations, and inspection scope, back into steps the company had treated as outside scope. Both outcomes trace to the same root cause — a designation built on the structural-fragment test alone, without the rest of ICH Q11's reasoning behind it.

None of this is exotic science — it is a documentation and process-mapping discipline that has to happen before the filing is drafted, not while a reviewer's question is sitting in an information request. Companies that treat the starting material boundary as a CMC strategy decision, backed by the full synthetic route and an honest impurity assessment, are the ones whose designations survive review the first time. Companies that lean on the structural-fragment sentence alone are the ones amending their process description mid-cycle.