8 Body-worn camera storage is the number that surprises agencies most after a camera rollout. The cameras themselves are a one-time purchase, but the video they produce keeps arriving every shift, and a meaningful share of it has to be kept for years. Getting the estimate roughly right before signing a contract avoids the two classic outcomes: running out of capacity mid-year, or paying for far more than the program ever uses. This article walks through the inputs that drive body camera storage, a simple model any records or IT manager can build in a spreadsheet, and the policy levers that change the answer far more than camera resolution does. The four inputs that drive body-worn camera storage Every estimate comes down to four variables: Bitrate, which depends on resolution, frame rate and codec. Recorded hours per officer per shift, which depends on recording policy and call volume. Number of recording officers and shifts per year. Retention, meaning how long each category of footage is kept before deletion. The first three determine how much new footage arrives. Retention determines how much of it accumulates. Gigabytes per hour by resolution Commonly cited industry figures put standard-definition body camera video at roughly 0.8 GB per hour, 720p at around 3 GB per hour and 1080p at roughly 12 GB per hour. Newer cameras using H.265 compression can produce noticeably smaller files at the same resolution than older H.264 devices, and some agencies record at lower resolution by default and switch to higher quality only for specific events. Because these figures vary by manufacturer and settings, the most reliable approach is to pull actual file sizes from a pilot group. A week of real uploads from twenty officers will tell you more than any spec sheet. How much footage an officer actually records Cameras rarely record a full shift. Most policies require recording during enforcement contacts and calls for service, with a short pre-event buffer. Industry guidance often uses an average of about 1 GB per officer per day once real recording behavior is factored in, which works out to roughly 250 GB per officer per year of new footage at standard settings. Higher-resolution settings, longer pre-event buffers and always-on policies can multiply that several times over. Patrol officers in busy urban precincts record far more than detectives or administrative staff, so segmenting the force by role produces a better estimate than a single average. A worked example Consider an agency with 300 officers wearing cameras, working about 230 shifts per year each, recording an average of 1.5 hours per shift at roughly 3 GB per hour. Daily new footage per officer: 1.5 hours x 3 GB = about 4.5 GB. Annual new footage per officer: 4.5 GB x 230 shifts = about 1 TB. Annual new footage for the agency: roughly 300 TB. Now apply retention. If 80 percent of footage is non-evidentiary and deleted after 90 days, and the remaining 20 percent is evidentiary and kept for an average of three years, steady-state storage is roughly: Non-evidentiary: 240 TB per year x (90/365) = about 60 TB on hand at any time. Evidentiary: 60 TB per year x 3 years = about 180 TB. Total: roughly 240 TB before any second copy, plus growth from long-held felony and homicide evidence that may never age out. Add a replica at a second site and the agency is planning for roughly half a petabyte within a few years. The general method is covered in storage capacity planning. Retention is the biggest lever on body-worn camera storage State law sets the floor. California’s Penal Code 832.18, for example, recommends keeping non-evidentiary body camera recordings for a minimum of 60 days and certain evidentiary recordings (such as use of force, arrests and complaints) for a minimum of two years, with access and deletion logs kept permanently. Other states set different minimums, and many agencies keep footage longer than required to cover civil claims. Small changes in policy have large effects. Extending non-evidentiary retention from 60 to 180 days roughly triples that portion of storage. Categorizing footage accurately at upload, so that routine recordings do not inherit the longest retention period by default, is one of the most effective cost controls available. Our article on digital evidence retention covers the policy side in more depth, and the general framework is in data retention policy best practices. Hidden multipliers Several factors quietly inflate body camera storage: Redacted copies produced for public records requests and court. Discovery exports duplicated for prosecutors and defense. Multiple angles when several officers record the same incident. Legal holds that freeze deletion for footage tied to litigation. Backups and replicas, which double or triple raw capacity depending on design. On the storage platform itself, protection overhead matters too. Erasure coding typically consumes less raw capacity than keeping full replicas for the same durability, as explained in erasure coding vs replication. Where the footage should live Recent footage is accessed often, for report writing, supervisor review and early case work. Older footage is rarely touched but must remain retrievable. That access pattern suits tiering: keep the first weeks or months on fast storage close to investigators and move the rest to a lower-cost tier that still responds in minutes rather than days. The concepts are explained in hot storage vs cold storage. Because retained footage keeps growing, the storage platform should scale in small increments without forcing a migration of years of evidence. Write-once retention locks help protect footage from deletion before its retention period ends. Our hub on digital evidence storage covers the full set of requirements, and cloud DEMS vs on-prem evidence storage compares hosting models. Planning for change: new cameras and new sources A storage estimate is only valid for the camera fleet and policy in place today. Most agencies refresh body cameras every few years, and new models often default to higher resolution, longer pre-event buffers or additional features such as automatic activation when a holster is drawn or a light bar turns on. Each of those can raise recorded hours or file sizes without anyone changing written policy. Body camera footage also rarely stays alone. Once an evidence repository exists, in-car video, interview rooms, drones, fixed cameras and video submitted by the public tend to flow into it. A sensible plan sets aside headroom for at least one new source and one camera refresh within the planning window, and treats the annual model review as the moment to check whether actual growth is tracking the forecast. Cost modeling tips Model costs per terabyte retained, not just per camera, so growth in retention shows up clearly. Include egress and export costs if footage will be stored in a public cloud; large discovery requests can add up (see egress vs ingress). Project at least five years, since evidentiary footage accumulates. Compare per-officer subscription pricing against capacity pricing at the agency’s projected volume, not today’s. Factor in staff time for categorization, redaction and deletion review. Checklist: sizing body-worn camera storage Measure real file sizes from a pilot instead of relying on spec sheets. Segment officers by role and recording frequency. Map every footage category to a retention period grounded in state law and agency policy. Account for redaction, discovery copies and legal holds. Decide how many copies you will keep and where. Plan for growth from long-held evidence that never ages out. Confirm the storage platform can expand without disruptive migrations. Revisit the model every year as recording policies and camera models change. Putting it together Body-worn camera storage is driven far more by retention policy and recording behavior than by camera resolution alone. Agencies that measure real footage, categorize it accurately and plan for evidence that accumulates over many years avoid the capacity crunch that hits many programs in their second or third year. Build the model once, revisit it annually and treat storage as a long-term part of the program rather than an accessory to the cameras. Frequently asked questions How many gigabytes per hour does a body camera record? Roughly 0.8 GB at standard definition, about 3 GB at 720p and roughly 12 GB at 1080p, according to commonly cited industry figures. Actual sizes depend on the camera, codec and settings. How much storage does one officer use per year? At standard settings and typical recording behavior, around 250 GB of new footage per year is a common rule of thumb. Higher resolution and longer recording can push that well past 1 TB. Does higher resolution always mean more storage? Generally yes, but modern codecs such as H.265 reduce file sizes at a given resolution. Retention policy usually has a bigger effect on total storage than resolution. How long must body camera footage be kept? It varies by state and footage category. California, for example, recommends at least 60 days for non-evidentiary video and at least two years for certain evidentiary recordings. Should body camera footage be stored on premises or in the cloud? Both are common. The decision depends on long-term cost at your projected volume, CJIS requirements, exit costs and how much control the agency wants over where evidence lives. Further reading Digital evidence storage for law enforcement Digital evidence retention Chain of custody for digital evidence CJIS compliant storage Storage capacity planning
