Studies from this time period were inconclusive and contradictory, mainly because of a failure to control simultaneously for other possibly influential variables. With data from Texas, analysis of variance (ANOVA) was used to determine whether offenders' race influenced the length of prison sentences for the crime of murder during 1923-72. In order to control for many variables simultaneously while ascertaining the influence of race on sentence length, a multivariate statistical technique was needed that could accommodate categorical independent variables and a continuous dependent variable. ANOVA was chosen as the most appropriate test to use in determining whether group means on a dependent variable are equal when the dependent variable is interval level, and one or more categorical variables, or factors, define the groups. Controlling for a number of variables using this method, offender's race apparently affected prison sentence, with black offenders receiving more lenient sentences during the 50-year period. The effects, however, are not invariant throughout the time period studied. During the 1920s and 1930s, African-American murderers were sentenced to longer periods of incarceration than whites; sentences equalized in the 1940s, and whites received longer sentences in the 1950s through the early 1970s. 2 tables and 34 references