We test several underlying assumptions regarding the impact of a concealed-handgun policy on violent crime rates. We posit that significant difference exist within geographic areas, and that permit holders reside in areas not prone to high levels of violent crime. We utilize aggregate-level data at the zip code level for Dallas, Texas, along with individual-level data on permit holders, a type of data which is used for the first time. We find stark differences across zip codes regarding the number of permits, sociodemographic characteristics, and violent crime rates. Permit holders are overwhelmingly white males and reside in areas with little violent crime. Those areas with high violent crime rates are the least likely to also contain a high number of residents with concealed-handgun permits. Researchers should be knowledgeable about the distribution of permits within geographic areas. This distribution is vital to properly specify empirical models.