WhichPodcast Research · Updated 30 June 2026

Podcasts are getting shorter: the typical episode is now just 35 minutes long.

New analysis of 22,394 RSS-distributed podcasts in the WhichPodcast index reveals how dramatically episode length varies by genre, and how the long-form interview boom of the late 2010s has played out against the parallel rise of short daily news shows. We give each podcast one vote — not each episode — so bulletin shows that publish thousands of clips a year don't dominate the stats.

35 min

median show’s average episode

42 min

mean across all shows

20–54 min

middle 50% of shows

1h 13m

90th percentile show length

Key findings

1.The median show in our index averages 35 minutes per episode (each show counted once, regardless of how often it publishes).
2.The mean across shows is 42 minutes, slightly higher because long-form interview shows pull the average upward.
3.18.2% of shows average under 15 minutes per episode.
4.Only 0.9% of shows average longer than three hours.
5.Gaming is the longest major genre, with a 69-minute median.
6.Artificial Intelligence is the shortest, with a 22-minute median.
7.Median show length has fallen from 44 minutes in 2008 to 38 minutes in 2025.

Some genres are dramatically longer than others

The longest genre averages 69 minutes per episode — more than 3.1× the shortest. Listeners self-select for runtime as much as they do for topic.

One bar per Apple-Podcasts category with at least 30 shows in our index. Bar width = the median show's average episode length.

Longest-running genres

Shortest-running genres

Artificial Intelligence22 min
News22 min
Education25 min
Business29 min
Fiction30 min

View full table — all genres ranked by median show length

Genre	Shows	Median	Mean
Gaming	229	1h 9m	1h 14m
Self-Improvement	53	55 min	50 min
Comedy	677	55 min	56 min
Society & Culture	127	45 min	46 min
Sports	486	45 min	45 min
History	620	40 min	51 min
Interviews	737	37 min	41 min
Science	488	37 min	47 min
True Crime	415	36 min	42 min
Crypto & Web3	104	35 min	35 min
Technology	729	33 min	36 min
Storytelling	228	31 min	41 min
Fiction	199	30 min	41 min
Business	478	29 min	32 min
Education	593	25 min	30 min
News	454	22 min	28 min
Artificial Intelligence	418	22 min	25 min

Episodes are getting shorter

In 2008, the median show in our index averaged 44 minutes per episode. By 2025 that had fallen to 38 minutes per episode — a -14% shift. Podcast episodes have not become longer overall — the median episode has actually shortened since the late 2000s. The long-form interview boom created a visible upper tail, but the catalogue overall has shifted shorter, likely driven by news briefs, daily shows, and short-form feed strategies.

Median show's mean episode length (in minutes) for podcasts that published at least one episode in each year. Years before 2008 excluded for low sample sizes; the current calendar year is excluded because mid-year data is biased toward high-cadence shows.

The shape of the podcast catalogue

Most shows average between 20 and 54 minutes per episode — but the long right-hand tail is real. 73 minutes is the 90th percentile, and a small but visible tail of shows average over three hours per episode.

Each podcast in the WhichPodcast index contributes one value — its mean episode length — bucketed into 15-minute groups up to 3 hours.

Why podcasts may be getting shorter

The data can't prove why podcast length has fallen, but several industry shifts may help explain the pattern:

Daily news briefings have proliferated. Shows like NPR's hourly newscast, BBC Minute and AI-narrated briefings publish at high cadence with very short runtimes.
Short educational episodes are a recognised format now. Five-to-ten-minute “learn one thing” shows have become a viable category.
Radio and broadcast audio gets repackaged as podcasts. Sermon clips, station-promo segments, and broadcast bulletins all live as “podcasts” in the RSS catalogue and tend to be shorter than originally-produced podcast content.
Creator burnout and production efficiency. Shorter episodes are dramatically easier to produce on a consistent schedule, which matters for retention.
Listening contexts have shrunk. Commutes, workouts, dog walks and short routines all favour episodes that fit inside a fixed window.
Platforms reward frequency. Spotify and Apple charts factor recent publishing into discoverability, nudging creators toward more, shorter episodes.
Audience expectations have shifted. The rise of YouTube Shorts, TikTok and short-form video has normalised micro-content consumption, even spilling into audio.

These are interpretations of the trend, not tested causal claims. Each would merit a dedicated study to confirm.

Long-form podcasting is still alive — it's just not the norm

Although the typical show now averages 35 minutes per episode, the long-form tail remains significant. The 90th percentile show averages 1h 13m, and several major genres sit well above the overall median.

Gaming (1h 9m), Self-Improvement (55 min) and Comedy (55 min) all sit comfortably above the overall median.

This suggests long-form podcasting hasn't disappeared — it has concentrated in genres where personality, discussion, and depth are central to the product. The shift toward shorter episodes is a story about how the catalogue's centre of gravity has moved, not about long-form going away.

What this means for podcast creators

The data does not suggest there is one perfect podcast length. Instead, it suggests listeners have different runtime expectations depending on genre.

For creators, the useful benchmark is not “make every episode 35 minutes.” It is: compare yourself to your category. A comedy show averaging 50 minutes may feel normal. A business advice show at 50 minutes may feel heavy. A news show at 50 minutes may feel unusually long unless it is built around deep analysis.

Don't chase the overall average blindly. A single catalogue-wide number hides everything that actually matters to your listeners.
Benchmark against your genre. The by-genre chart above is the more useful reference point — match the median for your category before you start experimenting outside it.
Shorter episodes may help discovery and consistency. Easier to publish weekly without burnout, easier for a new listener to commit to a first episode.
Longer episodes can work when listeners expect depth, personality, or story. True crime, long-form interview and narrative shows have built loyalty around runtime — but the format has to earn the minutes.
The best length is the shortest version that still delivers the promised value. If a 50-minute episode could have been 30 without losing anything, your retention curve will tell you.

How this differs from older studies

Previous analyses often quoted median episode lengths around 36–39 minutes. Our primary figure (35 minutes) sits below that range. Three likely reasons:

Podcast-weighting rather than episode-weighting. Most older numbers pool all episodes and take the median of that pool, which gives daily bulletin shows (publishing thousands of clips a year) disproportionate weight. We give each podcast brand a single vote regardless of how often it publishes. For comparison: the episode-weighted median of our catalogue is just 30 minutes ( 5,603,053 episodes in scope), and the episode-weighted mean is 40 minutes.
Daily news and short-update feeds have proliferated since most older studies were published. Even under podcast-weighting they shift the distribution shorter because there are now many bulletin-style shows in the catalogue, each one contributing a short-mean vote.
Broader RSS-catalogue inclusion. Earlier studies often sampled chart-toppers or registered hosting platforms; our index pulls from a wider RSS net, which surfaces more bulletin and micro-format shows.

We hedge these as “likely” explanations rather than tested causal claims — each would merit a dedicated study to confirm.

Methodology

Source: 22,394podcasts in the WhichPodcast catalogue with at least one ingested episode, sourced from Apple Podcasts, Listen Notes and direct RSS feeds. We use `audio_length_sec` as published in each feed's enclosure metadata.
Weighting: one podcast = one vote: For each podcast we first compute its own mean episode length, then aggregate (median, mean, percentiles) across that per-podcast set. Each show contributes one value regardless of episode volume, so a daily-bulletin feed publishing 8,000 short episodes counts the same as a weekly long-form interview show publishing 50 episodes. This is the right shape for the question “what is the typical podcast like” — readers picture brands, not the published-episode firehose.
Filters: Episodes with missing duration are excluded. Episodes with durations ≤ 0 seconds or > 10 hours are excluded as obvious feed-metadata errors. We lead with the median across podcasts because the long-form interview tail (multi-hour shows) pulls the mean upwards even after podcast-weighting.
Genre attribution: Each podcast carries one or more Apple-Podcasts category IDs. A show tagged with both Business and Comedy contributes its single per-podcast mean to both genre averages. We require a minimum of 30 shows per genre before publishing stats to avoid noisy small samples.
Year attribution: For each (podcast, year) pair we compute that show's mean length for episodes published in that calendar year, then aggregate across podcasts within the year. Years with fewer than 50 publishing podcasts in our index are excluded — typically the very early years of podcasting and the current calendar year (which is always partial mid-way through).
Known limitations: YouTube-only podcasts are not in scope. Our index is built from RSS feeds.
Popular podcasts are over-represented. Catalogue is sourced from Apple top charts and Listen Notes, so high-listener-score shows dominate. This is the relevant sample for most editorial framing (“what are popular podcasts like”) but it is not a representative sample of every podcast that has ever existed.
English-language over-representation.
Self-reported durations. We trust the duration field as it appears in the RSS enclosure. We don't re-encode or measure the audio.
Update cadence: Page regenerates daily. Current data slice computed on 30 June 2026.

Cite this report

Free to cite, quote and embed. Recommended citation:

WhichPodcast (2026). How Long Is The Average Podcast Episode? Retrieved from https://whichpodcast.com/research/episode-length

Writing about podcast trends?

Need a custom slice — a specific genre, country, year, length band, or top-N show list? Email zaq@lengmedia.com and we'll usually turn around a custom cut within 48 hours.

All charts on this page are available as PNG on request and the underlying aggregated data can be shared under a CC-BY licence.

← All WhichPodcast research