Researching agents before you query
A poorly researched query list is the main reason good books don't find agents. Agents specialize, their interests shift, and their lists open and close. Spending a week building a sharp list of twenty-five well-chosen agents will out-perform a month of mass-querying two hundred. This guide covers the databases, the Manuscript Wish List ecosystem, and the red flags of bad actors.
Where to find agents
Four standard resources, all widely used by querying writers:
- QueryTracker. A free database of thousands of agents with filterable search by genre, with writer-reported response times and status. The paid tier adds analytics; the free tier is enough for most querying projects.
- Publishers Marketplace. A paid industry service with deal listings ("Publishers Lunch") that show what each agent has recently sold and to which editor at which imprint. The single best resource for understanding whether an agent is currently active in your category.
- Manuscript Wish List (MSWL). A free directory of agent and editor wish-list posts, where they describe specific kinds of projects they want to see. Searchable by genre, theme, and tag.
- Literary agency websites. Ultimately, every agent has an agency page with official submission guidelines. Guidelines on the agency's own site take precedence over any third-party database.
Two other resources that aren't databases but are important: Writer Beware, maintained by SFWA, publishes alerts on known bad actors — fee-charging "agents," vanity scams, and industry-adjacent fraud. Absolute Write's Water Cooler hosts a long-running "Bewares, Recommendations & Background Check" forum where writers compare experiences with specific agents.
Reading a Manuscript Wish List
An MSWL post reads like a personal ad from an agent. Read it the way it's written: as a specific description of what this agent, right now, wants to see. If an agent says "I want literary horror grounded in the domestic," take them at their word. Your domestic-horror novel belongs in their query box. Your epic military fantasy does not, even if the agent represents fantasy in general — the MSWL is a more current, more specific signal than the agency's blanket category list.
MSWL posts include tags, which let you search across many agents for the exact niche you're in. If five agents have tagged "literary horror" in the last six months, those five are your first-pass list.
What MSWL doesn't tell you: whether the agent has actually been selling books in the category lately. Cross-reference with Publishers Marketplace or with the agency's list of recent deals before putting time into a personalized query.
Filtering your list
After you have a long list — say fifty names — cut ruthlessly:
- Wrong category. Cut any agent who does not represent your category. "Represents commercial fiction" is not the same as "represents literary fiction." A romance specialist is not the right home for your literary novel, no matter how lovely their website is.
- Closed to queries. Cut any agent whose submissions are currently closed. Save them for a future round.
- No recent deals. An agent who has not announced a deal in eighteen months may be slowing down, leaving the business, or on medical leave. A current drought is not proof of decline, but it is a flag worth noting.
- Obvious mismatch. An agent who sells almost exclusively romance and announces one literary novel every two years may sell literary novels, but your literary novel is not where you should start.
- Agency overlap. Most agencies have a "one query at a time" rule — if one agent passes, you can't immediately query another at the same agency. Respect this; if two agents at the same agency are equally good fits, query only one at a time.
A good working list is twenty-five to forty names, ranked roughly in order of fit. You'll query in small batches of six to ten at a time.
Vetting an individual agent
For each agent before you query them, answer five questions:
- Who is this agent's agency? Is it a full-service literary agency with editorial staff, or a solo operation?
- What have they sold recently, and to whom? Publishers Marketplace shows this. Their clients' acknowledgments often thank the agent by name.
- Do they have a track record in your category? Three deals in your category over the last two years is encouraging. Zero is a warning — not necessarily disqualifying, but requiring a reason.
- What do they actually want right now? MSWL, recent interviews, Twitter/Bluesky posts where they've mentioned looking for something specific.
- Are there any public complaints? Search their name on Absolute Write and Writer Beware. Check a handful of recent reviews on QueryTracker.
Red flags
A legitimate literary agent does not:
- Charge reading fees, editing fees, or submission fees. Agents are paid by commission on the deals they close. Anyone charging you up front is either not a real agent or is running a fee scam on the side.
- Require you to use a specific editing service before they'll consider representation. Some "agents" partner with pay-for-editing vanity schemes. Avoid.
- Refuse to disclose their client list or recent sales. Real agents are proud of their lists. Opacity is a flag.
- Promise representation before reading the manuscript. Representation is earned page-by-page. Anyone offering you a contract unseen is not operating in good faith.
- Demand exclusivity on queries as a default. Some agents ask for an exclusive read once they've requested a full manuscript. That's acceptable — for a limited window, two to four weeks. A demand that your query be exclusive for months is not standard.
- Offer a contract without AAR membership or equivalent credibility. Association of American Literary Agents membership is not mandatory, and some respected agents aren't members, but it's a useful signal in combination with a public track record.
Query order
Don't save your "dream agents" for last. Also don't send to your dream list first. A reasonable approach is to query in three waves: a first wave of six to eight agents who are a solid-but-not-desperate fit, so you can pressure-test the query on real reads; a second wave incorporating any changes you made from wave-one feedback; and a third wave of your best-fit dream list once the query has been optimized. The mass-query-everyone approach burns your best agents on an unpolished version of the pitch.