In studying the discretionary decisionmaking of criminal justice actors, much of the knowledge regarding factors that influence the decisions to arrest by police officers has been based on police observational data collected by researchers who have conducted in-person observation of the police at work. However, police are suspicious of outsiders and are not used to having their decisions scrutinized by supervisors, let alone an observer who is also a stranger. To this end, qualitative and quantitative data from a large-scale observational study of police, Project on Policing Neighborhoods (POPN), examined the impact of an observer’s presence on patrol officer arrest behavior. Two types of descriptive data on reactivity were prospectively built into the research design: the assessment of “ride level reactivity” and “activities and encounters with reactivity.” A major theme identified in the study was that patrol officers expressed concerns about safety. Patrol officers who expressed concerns about safety were less likely to arrest. The effect of concerns about safety is not simply a function of the overt or perceived dangerousness of the encounter but also a function of the sex of the observer. Officers who expressed concerns about safety and were accompanied by female observers were less likely to arrest. In addition, differences in findings between observational studies of the police conducted by male versus female field researchers may be partially attributable to the effects of observer sex and the patrol officer’s concerns for safety. The findings suggest that reactivity is a sex-specific and complex phenomenon because acclimation on the part of patrol officers over the course of fieldwork is not unidirectional. Study limitations and suggestions for future research are presented and discussed. References