No Response From Literary Agents? Query Problem or Manuscript Problem?

Silence from literary agents doesn’t come with a reason attached.

You send fifteen queries. Three ask for pages. None of them turn into an offer of representation, and the rest never reply at all. It’s tempting to read that pattern as a verdict on the manuscript. Sometimes it is. Often it isn’t — the actual cause is narrower and more fixable than a full rewrite, and it depends heavily on where in the process the silence happened.

Before you touch the manuscript again, it’s worth finding out which layer of the submission actually stalled. An agent who never requested pages is telling you something different than an agent who requested the full manuscript and then went quiet. Those two silences call for different next steps.

Before you diagnose anything, check what this agent actually received

Here’s a detail that changes the whole analysis: not every agency’s “query” is the same submission.

Some agencies genuinely want a query letter alone, with nothing else, and only ask for pages if they’re interested. Others build sample pages into the initial submission itself. Greene & Heaton’s guidelines ask fiction writers to send the first three chapters — roughly fifty pages — plus a synopsis, as part of the first submission. Eve White Literary Agency asks for the three opening chapters and a synopsis up front, the same way. If an agency’s initial submission already includes sample pages, you can’t safely assume from the outside that a “no request for more material” response was based on the query alone. The pages were part of the submitted package — that’s what the evidence supports. It doesn’t tell you how far the agent read, which component they weighed most, or whether the pages specifically caused the outcome.

That distinction matters more than which of the “three situations” people usually assume they’re in. Before interpreting any pattern of silence, the first useful question isn’t “query problem or manuscript problem” — it’s:

What material did this specific agent or agency actually have in front of them before they went quiet?

• A. Query or cover letter only, nothing else, no request afterward. The primary layers to check are targeting and the query/pitch — the sample pages genuinely haven’t been read yet, because this agency’s process doesn’t ask for them up front.
• B. Query plus sample or opening pages, submitted together as the initial package, no request afterward. The pages were part of the submitted package, so a no-request outcome can’t be attributed to the query alone from the outside. The primary layers to check now include the opening/sample pages alongside targeting and the query — not because the pages were definitely read in full or definitely caused the outcome, but because you can’t rule them out.
• C. An additional partial or fuller sample was requested after the initial submission, then a stall or decline. Interpret only the material that was actually reviewed at that stage, and consider whether the newly-requested pages changed the reading experience the earlier pages had created.
• D. The full manuscript was requested, then a stall or decline. The honest inference is that whatever material this agent saw before that request cleared their earlier screening threshold — not that a fixed, universal sequence of gates was passed, since agencies structure that sequence differently. After ruling out timing and agency policy, manuscript-level fit and execution become more relevant here than at the earlier stages.
• E. A full-request stall or decline recurs across multiple independent, well-targeted full-manuscript requests. A recurring pattern can justify a closer look at the manuscript. It still isn’t proof of structural failure on its own — pattern is not diagnosis.

None of these, by itself, proves the manuscript has failed. What changes is which layer is worth checking first, and that depends on knowing exactly what this agent already had.

Evidence-first query triage

Once you know what stage you’re actually in, work through the relevant layers below — cheapest and fastest to check first, most expensive to fix last.

1. Response policy and elapsed timing

Signal: No response yet, at any stage.

What to check: Each agency publishes its own current response policy and stated timeframe. Some operate on a “no response means no” policy for queries specifically, which they disclose in their own submission guidelines — meaning silence there isn’t a signal about the manuscript at all, it’s simply how that agency declines.

What not to infer: That silence inside a stated window means anything about quality. An unfinished queue is not a verdict.

What evidence would change the diagnosis: The agency’s current submissions page states a window that has clearly and fully elapsed, or explicitly confirms they only reply to yes.

Next smallest action: Re-read each target agency’s current submissions page and note their specific stated policy — don’t rely on general advice about what agents “usually” do, since guidelines change and only the current published version is authoritative.

2. Target-agent fit

Signal: Queries are going out and getting no requests, even after the response window has passed.

What to check: Whether each agent currently represents your genre and category, and what they’ve recently sold or said they’re actively seeking. Agent wish lists shift; a list built a year ago may no longer reflect who’s acquiring what you’ve written.

What not to infer: That a stalled, genre-mismatched query is evidence the manuscript isn’t ready. It’s evidence of a targeting miss.

What evidence would change the diagnosis: A well-targeted, correctly-timed query to an agent who actively represents your category still produces no request.

Next smallest action: Re-verify each agent’s current representation focus against their own current bio or agency page, not a list compiled months ago.

3. Query and submission package

Signal: Targeting looks right, timing has genuinely elapsed, and requests still aren’t following.

What to check: Whether the query clearly conveys premise, stakes, and who the book is for. Follow each agency’s exact current requirements for metadata — category, genre, word count, and any other detail they ask for vary by agency, so confirm what a specific agent’s guidelines actually request rather than assuming a standard list. Comparable titles are worth including only where an agency requests them or where they’re commercially useful, not as a universal requirement.

What not to infer: That a flat or unclear query proves the manuscript underneath it is weak. Plenty of strong books get undersold by an underwhelming pitch.

What evidence would change the diagnosis: A cold read of the query — by someone who has never seen the manuscript — doesn’t convey why this book, specifically, matters.

Next smallest action: Have someone outside the project read the query without context and describe back what they think the book is about and why they’d want to keep reading.

4. Opening or sample pages

Signal: Situation B or C above — sample pages were part of the initial submission, or were requested afterward — and the outcome consistently stalls or converts to declines.

What to check: Whether the opening pages deliver what the query promised. Many agencies build sample or opening pages directly into their submission process — either up front, as Greene & Heaton and Eve White both do, or as a follow-up request — specifically because those pages are a meaningful, separate evaluation layer from the query, not a formality. Voice, pacing, and how quickly the central tension appears all get tested here in a way the query alone doesn’t test.

What not to infer: That a stall after a sample or partial request means the whole manuscript needs structural work. It may mean only the opening chapters aren’t doing the job the query told the agent they would.

What evidence would change the diagnosis: The query’s promise and the opening pages’ actual execution diverge when read back to back.

Next smallest action: Compare your query’s promise against your opening pages, line by line, and isolate whether the opening specifically — not the whole manuscript — is the gap.

5. Full-manuscript pattern

Signal: Timing has passed, targeting was sound, the query reads well cold, and the pattern recurs across multiple independent, well-targeted full-manuscript requests.

What to check: Pacing, stakes, and structural execution across the manuscript as a whole, ideally through a fresh read from someone who hasn’t already read multiple drafts of it.

What not to infer: That one data point — one agent, one quiet response — is conclusive. A pattern across a meaningful sample is what’s informative, not a single instance.

What evidence would change the diagnosis: The same kind of concern (pacing collapse, unclear stakes, structural confusion) turns up independently across more than one full-manuscript read.

Next smallest action: Get a structured, second-opinion read focused specifically on where interest seems to be dropping, rather than a general “does this work” pass.

Check the agency’s current AI-use rules before you submit anything

This part isn’t optional, and it isn’t a verdict on whether AI tools have any place in a writer’s process. It’s about matching your submission to the specific, current rules of the agency you’re targeting.

Agency policies on this differ, and they’re evolving. Some restrict AI-generated material specifically. P.S. Literary Agency states they do not represent work generated by artificial intelligence, citing unresolved questions around copyright and authorship — a policy focused on AI-generated material.

Others go further and restrict AI-edited material too. Greene & Heaton’s submission guidelines state that submissions originated, written, or edited using AI will not be accepted — explicitly covering the cover letter, synopsis or proposal, and manuscript. Eve White Literary Agency goes further still. Their submissions page states they do not accept cover letters, proposals, synopses, or manuscripts written using generative AI, and that all submissions are read by their team without AI involvement — and their FAQ is explicit that they “cannot currently accept any work that has been generated or edited using AI,” matching Greene & Heaton’s broader restriction rather than the narrower generated-only policies above.

Other agencies haven’t published a position at all. That range — from no stated rule, to restricting generated material, to restricting generated or edited material — is exactly why a general rule isn’t useful here. What matters is the specific, current guidance published by each agency you’re targeting, checked at the time you submit, not a rule of thumb copied from somewhere else.

Before you query or resubmit: re-read each target agency’s current submission page for language about AI-generated or AI-edited material, and make sure your submission complies with what that specific agency has published today.

Diagnostic summary table

What the agent had	What happened next	What that evidence supports	What it does not prove	Next smallest check
Query/cover letter only	No response yet, inside the stated window	Nothing — the queue is simply unfinished	Anything about the manuscript or the query	The agency’s current published response policy
Query/cover letter only	No response, window elapsed	The query alone didn’t earn a next step	That the sample pages were ever read — they weren’t submitted	Target-agent fit, then query/package clarity
Query + sample/opening pages (initial package), window elapsed	No response	The pages were part of the submitted package — a no-request outcome can’t be attributed to the query alone from the outside	That the query was solely responsible, or that the pages were definitely read in full or caused the outcome	Target-agent fit, query clarity, and whether the opening pages deliver the query’s promise
Additional partial/sample requested after initial submission	Stall or decline	The earlier material cleared a real gate; the newly-requested pages are now the more relevant layer	That the whole manuscript needs a structural rewrite	Whether the newly-requested pages sustain what the earlier material promised
Full manuscript requested	Stall or decline	Whatever material this agent saw beforehand cleared their earlier screening threshold	That a universal, fixed sequence of gates was passed — agencies structure this differently	Manuscript-level pacing, stakes, and structure — after timing/policy is ruled out
Full-request stall/decline repeats across several agents	—	A repeated pattern carries more weight than a single instance	That the pattern alone tells you exactly what to fix without a closer look	A fresh, structured read focused on where interest drops

Frequently asked questions

Does no response always mean rejection?
Not necessarily, and not right away. Some agencies use a “no response means no” policy after their stated window has passed, which they disclose in their own guidelines. Inside that window, no response is an unfinished queue, not a decision.

If I get full requests but no offers, does that mean the manuscript needs a rewrite?
Not automatically. A full-manuscript stall is more informative than an earlier-stage stall, because whatever material the agent saw beforehand cleared their earlier screening threshold — but one declined full request is still one data point, and agencies structure their earlier stages differently. The more useful next step is a fresh, structured read before assuming the largest, most expensive fix.

How many silent queries before I should suspect the manuscript instead of the query letter?
There’s no fixed number, and no verified figure to cite here — this article intentionally avoids stating a specific threshold that isn’t backed by real data. What’s informative is a consistent pattern across a meaningful sample of well-targeted, correctly-timed queries, not one or two responses.

Should I revise my query or my manuscript first?
Generally, check and fix the cheaper, faster layers first: confirm timing and targeting, then test the query, then the opening pages. Move to full manuscript-level work only if the pattern persists after those are ruled out.

Is it ever safe to use AI tools on my submission materials?
That depends entirely on the specific agency you’re submitting to. Some, like P.S. Literary Agency, currently restrict AI-generated material. Others, like Greene & Heaton and Eve White Literary Agency, restrict AI-generated or AI-edited material. Others haven’t published a position. Check each target agency’s current guidelines before you submit, every time.

What to do with the answer

If the pattern points to timing or targeting, the fix is research, not rewriting. If it points to the query or the opening pages, that’s a scoped, testable revision. If it consistently points deeper — into pacing, stakes, or structure across full-manuscript requests — that’s worth a second opinion before another round of querying or another round of paid editing on the wrong layer.

Be My Tech’s Manuscript & Publishing Readiness Diagnostic exists for that last case: a fit-reviewed, evidence-based read on what a manuscript actually needs next, before you commit to a direction — including when the honest answer is that no further diagnostic is needed yet.

Sources

• Submission Guidelines — P.S. Literary Agency
• Submission Guidelines — Greene & Heaton
• Submissions — Eve White Literary Agency
• Eve White Literary Agency — FAQ (AI policy)
• Literary agents urge writers to avoid AI as they see “change in nature of submissions” — The Bookseller

Related posts

• Should You Query Literary Agents or Self-Publish? A 2026 Manuscript-Readiness Framework — this article assumes querying is the chosen path and diagnoses why it’s stalling; that article addresses the earlier route decision.
• Authors & Books hub — primary Author landing page and intake.