Home / Reports / Measured ride downtime at North American theme parks, summer 2026

Measured ride downtime at North American theme parks, summer 2026

Wait times tell you how long the line was. They do not tell you whether the ride was running at all. Thoosie records an explicit operating status for every ride at every park it tracks, about once a minute, so it can measure the second thing.

Between 2026-05-23 and 2026-08-15, across 1,263 rides at 34 North American parks, Thoosie recorded 44,744 separate unplanned outages totalling 44,749 hours out of service, measured against 921,020 hours of observed in-service time. Mean availability across rides was 94.8% (an unweighted average of per-ride availability). Scheduled refurbishment is recorded separately and is not counted as an outage.

The rides that were down the most

Availability is the share of a ride's scheduled in-service time during which it was actually running. Time when the ride was closed for the day or the season is not in the denominator. Only rides with at least 300 recorded in-service hours are listed, so every figure rests on real exposure rather than a thin sample.

Breaking often is not the same as being down long

Some rides stop frequently and restart quickly. Others fail rarely and stay dark for hours. Ranked by number of separate stops rather than time lost:

How this was measured

Thoosie polls the public ride-status feed each park publishes, roughly once a minute, and records the status it reports: operating, closed, down, or under refurbishment. An outage is a continuous run of down status inside the park's published operating hours. Stoppages where 60% or more of a park's in-service rides were down at the same moment are excluded, because those describe weather or a park-wide event rather than one ride failing: that removed 16% of events and 19% of recorded downtime. Days a ride never ran at all are also excluded, so a seasonal closure is not counted as a breakdown.

What this report deliberately does not say

It does not rank parks. Feeds differ in how readily they report a ride as down, and those differences are large enough to swamp any real difference in maintenance. One park in this dataset logs 129 outages across 51 rides while another logs 2,959 across 44. That gap describes reporting behaviour at least as much as it describes reliability, so a "most reliable park" ranking built on it would be misleading.

For the same reason, eleven parks are excluded entirely: their feeds never report any ride as down, which would render them as flawless. They are not included in any figure above.

Using this data

Free for journalistic, academic and personal use with attribution to Thoosie and a link to the page you took the figure from. Every ride named above links to its own page, where the same numbers appear alongside the measurement window. Custom cuts, any park or date range, on request. Method: how we measure.