This study addressed two questions believed to have implications for the current definition of mild mental retardation. The sample of 60 males (30 whites and 30 blacks) was selected from a population that had been evaluated according to the dual criteria of mild mental retardation in the public schools of southern Louisiana. Measures of the students' IQ, adaptive behavior, and socioeconomic status were obtained from the files. The Digit Span subtest of the WISC-R, a serial learning task, a paired-associates learning task and the Porteus Mazes were individually administered to the subjects. Other measures obtained were the regular homeroom and special education teachers' ratings of the students' social skills in the school setting;The results revealed that those students who had higher adaptive behavior scores did not perform better on the Level I tasks. Likewise, social competence as measured by adaptive behavior did not correlate to any noticeable degree with teachers' ratings of social skills in the school setting. However, better teacher ratings of social skills were associated with better performance on the Level I tasks. IQ and social competence as measured by adaptive behavior and teachers' ratings of social skills were not discrepant in the predominantly lower-class sample. The correlations of Race and Socioeconomic status were virtually zero with IQ and adaptive behavior. Likewise, race and Socioeconomic status were not significantly correlated with performance on the Level I tasks and the teachers' ratings of social skills. Although IQ correlated better than the adaptive behavior measure with the measures of Level I abilities and the teachers' ratings of social skills, a linear composite of the variables of IQ, Race, Adaptive Behavior, and Socioeconomic status was poorly predictive of the performance of the students.