The present study investigated the reliability and validity of scores from the Child Abuse Potential (CAP) Inventory. The CAP is a screening device for physical abuse potential in adults; it is a self-administered test written on a third-grade reading level. Subjects were 113 mothers, including 53 (46.9%) mothers of young children with handicaps, and 60 (53.1%) mothers of young children without identified handicaps. Analyses of CAP validity scales suggested that the Random Response subscale could be improved by omitting selected items. Although total CAP scores had an impressive alpha coefficient (0.91), consistent with those reported in previous research, some short subscales had unacceptable coefficients. The factor structure underlying responses was interpretable, but different in some respects from results reported in previous studies. One table presents the varimax rotated solution of the CAP, and three appendices list details of the factor analysis. A 13-item list of references is included. (Author/SLD)