Virtually all of the empirical inquiry on police-involved shootings has focused on only one of the two avenues for optimizing the frequency of police use of deadly force: shrinking the outer limits of officer discretion to shoot, primarily through adoption of restrictive departmental policies and more stringent review of shootings. Future research should be conducted on the second avenue: officer restraint in the face of authority to shoot. Sizable data sets on high risk police-civilian encounters should be developed because some of the useful police violence reduction techniques may not be obvious and others may be counterintuitive. Methods for systematically studying police officer restraint in the use of deadly force should also be established. Successes should be studied rather than failures. Among the archival sources that might prove useful are case reports and firearms use reports. Additionally, officers could be surveyed and those involved in high-risk encounters could be interviewed. Personal observations could be conducted, jail or State correctional facility inmates could be surveyed, and reports by police department supervisors could be reviewed. Possible variables to be investigated include type of assignment, type of opponent, and lighting conditions. Results could provide systematic knowledge, grounded in actual police encounters, about high-risk, police-civilian encounters. Approximately 250 references are included.