All 17 State institutions were included in the study. Contractual, prerelease facilities not employing CO's were excluded. The study encompassed 2,713 CO's. A baseline, self-administered prequestionnaire examined the importance of selected CO variables in predicting officer involvement in inmate assaults. Data on the occurrence and content of battery cases were obtained from institutional records. Sociodemographic information for the CO cohort was secured from personnel files. The obtaining of the daily inmate census permitted an examination of the relationship between prison population size and inmate batteries on CO's. Incidence rates for battery and injury of CO's were 65.3 and 16.8 per 100,000 work hours, respectively. Officer age, race, sex, rank, length of employment, type of institution, height, weight, past history of battery injury, coping strategies, and behavioral intentions relevant to conflict with inmates were all predictive of involvement in battery episodes. Tabular data, study instruments, and 200-item bibliography.