In view of the apparent impossibility of predicting any isolated act of violence, the study approach involved a search for some dimension on which people may be ranked so that those who have committed criminal violence are thereby identified. The hypothesized dimension is coping capacity, and the variables for determining individual probabilities were personality traits, attitudes, and social perceptions. Age, intelligence, sex, ethnic group, and socioeconomic background were controlled insofar as possible by appropriate matching of subjects. Midsentence inmates volunteers were paid to complete the Birkman Method. About 2,300 inmates responded. From this pool, all of the following categories were retained: 92 women, 142 Hispanics, 270 inmates who had records of having threatened a violent act, and 461 inmates who had been convicted of committing one or more violent acts. The psychological test results of these subjects were evaluated. For comparison purposes, a sample of working people holding jobs of the type and at the level that parolees and ex-offenders seek and obtain was taken from a large, privately owned data base. A multiple regression predictive equation for violence developed solely from two-thirds of the inmate sample cross-validated significantly with the randomly withheld remaining third of each category. However, it grossly overpredicted violence among semiskilled workers. The study identified three major aspects of the coping dimension. Two of these aspects were neuroticism and extraversion. The third major aspect of the dimension of coping capacity involved attitudes and values. The violent inmates had almost twice the tendency of the adequate workers to answer the personality questions in contrary, unpopular directions. Two tables, 5 footnotes, and 35 references are provided.