This study aims to assess the effect of body-worn cameras (BWCs) on police use of force, in a British police force context. We tested the effect of BWCs with a large British force in a six-month randomised controlled trial. Police shifts (n = 430) were randomly assigned on a weekly basis into treatment and control conditions. Odds ratios of use-of-force rates per arrests were used to estimate the causal impact of BWCs. Analyses of these odds for overall use of force and again within pre-specified force categories were conducted. Overall, we found a 50 percent reduction in the odds of force used when BWCs are present compared with control conditions. Our estimates suggest a 35 percent reduction of overall weighted force in the treatment conditions compared with control conditions. However, the effect concentrates in open-hand tactics (physical restraints and non-compliant handcuffing), with no discernible effect on categories of more aggressive force responses (for example, dogs, Tasers, batons, pepper spray); 40 percent ‘more force’ was detected in treatment conditions for handcuffing non-combatant suspects. We conclude that BWCs deter officers, offenders or both into complaint behaviour. Importantly, showing a conditional effect on force types can be further contextualised as enhanced transparency and accountability by the police, with greater reporting of use of force that would otherwise be concealed. Our findings illustrate the importance of analysing police use of force with and without compliant handcuffing of arrestees, which may or may not form part of the force continuum.