Small, inexpensive, often poorly made handguns known as "Saturday night specials" are disproportionately involved in crime. Maryland banned the sale of Saturday night specials effective January 1, 1990. During the 2 years between the law's passage in 1988 and its effective date, legal handgun sales in Maryland were 34% higher than expected (p = 0.09). Interrupted time-series analysis of age-adjusted homicide rates for 1975-1998 with statistical controls for trends in two neighboring states, social and economic variables, and temporal patterns in Maryland's homicide rates was used to assess the effect of the law. Estimates of the Saturday night special ban effects depended on the assumption made about the timing of the law's effects. Models that assumed a delayed or gradual effect of the Saturday night special ban produced estimates indicating that firearm homicide rates were 6.8-11.5% lower than would have been expected without the Saturday night special ban (p < or =0.05). The model that assumed an immediate, constant change in response to the law showed no law effect, unless an outlier was excluded from the analysis. Excluding this outlier, the model estimated a 15% increase in firearm homicides associated with the Saturday night special ban. None of the models revealed significant law effects on nonfirearm homicides.