There is a considerable body of evidence which indicates that the presence and/or density of on-premise alcohol outlets (e.g., bars, restaurants) has a positive association with violent crime. But the empirical evidence for a causal relationship is weak. This can in part be attributable to a lack of temporal component in existing analyses. Whether violence occurs in and around bars due to an alcohol-induced lack of self-control in patrons, or due to a temporary increase in the ambient population, these effects should only be present during active times of day (i.e., opening hours). If the likelihood of violence occurring on a street (or nearby streets) increases only during bar opening hours, or shortly after closing, then we would have stronger evidence for a causal relationship. To disentangle the causal link, this study uses spatially-sensitive data on emergency calls for service and corresponding information on the opening times of bars in two major Dutch cities.