Seventeen extrafamilial homicidal child molesters (HCM) were compared to 35 convicted extrafamilial child molesters (CM) who had not murdered, or attempted to murder, their victims. The two groups did not differ on age, IQ, education, history of marriage, or family history, although marriage rates of both groups were well below the national average. Similarly, both groups had high rates of features representing family instability. HCM more frequently victimized strangers. The results on the self-report psychological inventories, the Derogatis Sexual Functioning Inventory (DSFI) and the Buss–Durkee Hostility Inventory (BDHI), did not distinguish between the groups, although the DSFI revealed sexual inadequacy in both groups. The BDHI did not describe the groups as pathological. The Psychopathy Checklist—Revised Total Score, Factor 1, and Factor 2 described both groups as demonstrating high levels of psychopathy, with the HCM scoring significantly higher. Factor 1 and Factor 2 scores placed the HCM group in the 93rd and 82nd percentiles, respectively, compared to published norms for forensic patients. A greater proportion of HCM suffered from antisocial personality disorders and paraphilias, especially sexual sadism. Over 53% of the HCM, and none of the CM, were comorbid for pedophilia and sexual sadism. Significantly more HCM received three or more DSM III-R diagnoses. The phallometric assessments generally supported DSM diagnoses. The HCM demonstrated significantly higher levels of deviant arousal to pedophilic and adult assault stimuli. Police files revealed that, prior to the index offense, a significantly greater proportion of HCM had been charged with, or convicted of, violent nonsexual, and sexual offenses. The HCM had been charged with, or convicted of, more than 2.5 times as many criminal charges than the CM. A discriminant function analysis revealed that two variables, Factor 1 of the PCL-R and the number of violent entries in the police records, correctly predicted 78.6% of the HCM and 97.1% of the CM. Implications for understanding homicidal child molesters and for future research are discussed.