Methodology

How We Assess Contest Fees

Some rows in our contest database carry small indicators about fees, prizes and deadlines. This page documents exactly how each one is computed, so you can check our arithmetic. All of it is mechanical: the indicators are derived from the listing's own stated fee and prize, never from an opinion about the organiser.

What these indicators are not

None of these indicators is an accusation. A high entry fee relative to the prize is a fact about two numbers on the organiser's own page. Plenty of legitimate contests charge meaningful fees, because fees fund the prize pot, pay judges and keep magazines alive. The indicators exist so you can weigh cost against potential return before you enter, not to tell you a contest is good or bad.

The indicators, one by one

1. Free entry

Shown when the listing is marked free-entry or states a zero fee. 47 of the 110 contests we track (42.7%) are free to enter. Free contests have no fee-to-prize arithmetic: whatever the prize, your downside is your time.

2. Cash prize listed

Recorded when the listing's prize includes cash, as opposed to publication only, tuition waivers, mentoring or goods.101 of 110 listings (91.8%) list a cash prize.

3. Prize relative to fee

Where a listing has a paid, parseable entry fee and a parseable cash prize in the same currency, we divide the headline prize by the primary entry fee. Both numbers come from the organiser's own listing:

Primary fee. The first amount in the fee text. "£8 first poem, £6 after" counts as £8.
Headline prize. The first amount in the prize text. "£1,000 first, £300 second" counts as £1,000.
Same currency only. We never convert currencies, so a euro fee against a dollar prize produces no ratio.
Unparseable rows produce nothing. "See website" or "cash award" rows are excluded and counted as unparsed, not guessed.

That yields a ratio for 37 of the 110 listings. The current distribution: median125x, middle half between 56.1x and 227.3x, full range20x to 625x. The full distribution is charted in theState of Writing Contests 2026 report.

The "high fee relative to prize" threshold

A listing carries the note high fee relative to prize when its ratio falls in the bottom quartile of the distribution above. The threshold is not hand-picked: it is the lower quartile of the computable ratios, recomputed every time the statistics regenerate. As of 20 July 2026 that threshold is56.1x, and 9 of the 37computable listings (110 tracked in total) fall below it.

Concretely: at a 56.1x threshold, a contest charging a £10 fee sits below it when its headline prize is under £561. Because the threshold is a quartile of the live data, it moves as the database grows; the note always means "in the bottom quarter of the contests we can measure", not "above a magic number".

4. Deadline confidence

Listings show deadline to confirm when the date is projected from a previous cycle rather than read from the organiser's page (67 of 110 currently), and a checked date when our weekly verifier last confirmed the record against its source page (30 of110 carry one). Records whose source link fails verification are suppressed from the public lists entirely rather than badged.

How to use this when choosing contests

Treat entry fees as a submissions budget, the way freelancers treat postage: decide what a contest cycle is worth to you per year, and spend it on the strongest fee-to-prize positions for your genre.
A modest fee with a modest prize can still be worth it when the prize includes publication, an agent read or a course. The ratio measures cash against cash; it cannot see those.
Check what the fee funds. Fees that include a magazine subscription or written feedback buy more than a lottery ticket.
Prefer confirmed deadlines when planning close to a date; projected deadlines can move when the new cycle is announced.

Reproducibility

The parsing rules and thresholds on this page are implemented in one deterministic script (scripts/build-contest-stats.mjs in the site repository) that reads only our contest database. Every figure on this page and on the listings regenerates from it; nothing is adjusted by hand. Statistics last generated 20 July 2026 from 110 records.