Objective To examine the impact of a novel firearm ‘pointing’ policy that requires officers to document when they directly point their guns at citizens.

Methods Sixteen years (2003–2018) of narrative officer-involved shooting (OIS) reports from the Dallas Police Department were qualitatively coded to explore both the total frequency and specific characteristics of OIS before and after the policy change in 2013.

Results χ² tests found that the firearm pointing policy was associated with a reduction in the proportion of ‘threat perception failure’ shootings (ie, those where an officer mistakes an item for a gun). Auto Regressive Integrated Moving Average analysis found that the policy change was associated with a gradual, permanent reduction in total OIS; however, that impact was not immediate.

Conclusions Firearm pointing policies have the potential to alter organisational behaviour, particularly in highly discretionary shootings. It is unclear whether the specific mechanisms for the changes include more accountability through constrained discretion, reduced options to handle situations once officers’ guns are drawn and pointed, or an effect on officers’ timing and vision during ambiguous scenarios.

Policy implications Although organisational change may be a long and complex process, reductions in OIS can prevent serious injuries and death. The policy change did not lead to an increase in the proportion of officers injured during OIS incidents.