Eysenck has postulated a general theory of criminality that predicts that criminals as a group will have significantly higher scores on the personality dimensions of psychoticism, extroversion-introversion, and neuroticism-stability. Furthermore, the theory supposes that inherited features of the nervous system influence criminal behavior. Although the theory has been supported in studies of European offenders convicted of nonviolent crimes, it has not been sufficiently tested with American, and particularly with minority, offenders. Researchers tested this theory by administering the questionnaire to 398 American inmates, 248 blacks, 121 Hispanics, 25 whites, 1 Indian, and 1 inmate of mixed ethnic background. The sample corresponded to the ethnic composition of the prison population. A control group of 187 unemployed males resembled the criminal group in all demographic variables. The overall results failed to support Eysenck's predictions. A significant difference between the two groups was found in the L scale only. The L scale measures attempts to make oneself look socially desirable. No differences were found for psychoticism and neuroticism, and criminal subjects scored much lower on the extroversion-introversion scale. When the criminal sample was divided according to offense type, analysis of the scores still failed to find significant differences. When high L-scorers were eliminated from the sample (Eysenck warns against imparting too much significance to L scores), differences continued to be insignificant. Therefore, the Eysenckian theory when applied to American blacks and Hispanics appears to be invalid. Three references are provided.