Extant research on police use of force has established that force is not distributed evenly across communities. However, this body of research is limited compared to the attention afforded to the influence of individual and situational factors on police use of force. Further, findings on the influence of community characteristics have been inconsistent and the overwhelming majority of this research has been generated out of policing jurisdictions in the United States (U.S.), so the transferability of these findings to other jurisdictions, such as Australia, is unclear. There is also a dearth of rigorous translational research that can inform police agency responses that might reduce police use of force across communities. While community-oriented policing (COP) has been associated with reduced police use of force (theoretically and, recently, at the policy level), to date there is no empirical evidence to support the association. This thesis aims to address these critical evidence gaps: to explain variations in police use of force across communities by examining the influence of community characteristics and COP on police use of force in Australia. The thesis comprises four empirical studies to address the research aim. Study 1 aimed to determine the influence of community characteristics on the distribution of police use of force across communities, examining the most commonly applied theories and community characteristics used to explain community-level distribution of police use of force from the extant literature – minority threat theory, ecological contamination theory, social disorganisation and socio-economic disadvantage. This study utilised data from police use of force reports from a three-and-a-half-year period (comprising 18,322 police use of force events) in combination with recorded crime data and census data. Findings from a negative binomial regression analysis indicated support for ecological contamination theory (violent crime rates), social disorganisation and partial support for socio-economic disadvantage in predicting the frequency of serious police use of force events across communities, while minority threat theory was not supported. In contrast, results from a multilevel ordered logistic regression analysis suggested that individual and situational factors are more influential than community characteristics in predicting force severity. Study 2 examined the extent to which the distribution of citizen behaviours and risks across communities may be a driver for some of the variation in the frequency of police use of force across communities that were found in Study 1. This study drew on the same data sources as Study 1 and used Latent Class Analysis (LCA) to identify a typology of citizen behaviours in police use of force events, drawing on citizen behaviours and risks reported in police use of force reports. A five-class typology demonstrated best fit, comprising the following classes: ‘violent, unstable and drug or alcohol affected’; ‘apparent mental disorder and possessing a weapon’; ‘alcohol-related violence’; ‘suspicious and fleeing’; and ‘violent behaviour and threatening a weapon’. Additionally, these citizen presentation classes were associated with distinct offence types, contexts for use of force, and citizen characteristics. Multilevel logistic regression determined that the five citizen presentation classes were differentially distributed across communities, with three of the most prevalent classes showing the most significant associations with community characteristics. This indicates that the distribution of citizen behaviours and risk factors in police use of force events are also influenced at least to some extent by community factors. Study 3 examined whether a policing approach, specifically COP, may be able to impact on the frequency with which police and citizen encounters result in the use of force by police, focussing in particular on communities where the propensity for police use of force may be higher. This study drew on a sample of 64 socially challenged communities in an Australian policing jurisdiction. It examines the association between Officer-in-Charge (OIC) reports of formal and informal community consultation in a problem-solving context (elicited through an online survey), and the frequency of police use of force (drawn from police use of force reports). This study found that higher levels of COP were associated with lower rates of police use of force in high violent crime communities, but not low violent crime communities. The findings of this study support the contention that COP may reduce the use of coercive policing tactics. Study 4 further examined the impact of COP on police use of force, this time focussing on how higher levels of community engagement by officers may influence their attitudes towards coercive policing responses. Additionally, this study tested one pathway through which community engagement may influence police use of force: through the reduction of officer perceptions of social distance from the community. Study 4 drew on the same sub-sample of socially challenged communities in Australia as used in Study 3. Survey responses from 300 officers from 55 Divisions were used in the study, along with a number of other publicly available data sources. Regression analyses supported the contention that COP is associated with a reduced propensity for coercive policing at the officer level, and an increased propensity for non-coercive policing, aligning to the findings of Study 3. The findings also indicate that social distance mediates the relationship between officers’ community engagement and their endorsement of non-coercive policing responses but not coercive policing responses. Overall, the collective findings of this thesis suggest that community matters when it comes to police use of force in Australia. A number of characteristics of communities, such as violent crime, social disorganisation and socio-economic disadvantage, increase the frequency with which police use force. The influence of these community characteristics is likely to be driven by both tangible and intangible factors. In terms of tangible factors, these community conditions can drive concentrations of particular types of social problems, offending and disorderly behaviour, which may elicit more assertive or coercive policing responses. However, officers’ perception of social distance is an intangible factor that is also affected by community conditions and which is associated with the propensity to use non-coercive policing. COP approaches were found to be a policing approach that can reduce police use of force in high violent crime communities, at least in part by reducing officers’ experience of social distance from the community.