Although the possible effect of race on sentencing decisions is a much-studied question, even recent studies suffer from methodological problems. This paper attempts to correct these problems by using a large number of cases and a large number of offenses, by dividing the sentencing decision into two separate decisions, by using an appropriate scale to measure sentence severity, by including controls for relevant legal and extra-legal factors, and by using multivariate analysis. Our major findings are that race does not have a direct effect on sentence severity, but that blacks are more likely than whites to be incarcerated.