The sample included 1,944 male infants born consecutively from January 1, 1936, to September 30, 1938, at Rigshospitalet in Copenhagen, Denmark. Ultimately, 544 cases were dropped from the study because positive identification of the father was not possible. Of 1,400 fathers remaining, 92 or 6.6 percent were severely sanctioned, i.e., they had served at least one prison sentence. Therefore, the sons of the 92 fathers were considered to be at high risk for criminal involvement, and 39.1 percent had received at least one prison sentence by the time of the study. The risk of serious criminal behavior was 5.6 times greater among cohort members whose fathers had been severely sanctioned than among cohort members whose fathers had not committed any offense. The mean IQ for criminal boys was 102 versus 108.8 for noncriminal boys, a statistically significant difference. After closer examination of the data, however, it became clear that seriously criminal boys had a significantly lower IQ than noncriminal boys only in the high risk condition. The findings generally supported all three hypotheses, indicating that primary prevention of antisocial behavior may need to focus on making school a rewarding and bonding experience for all children in order to lessen criminal behavior tendencies. 21 references