The inconsistent results of studies assessing the relationship between broken homes and crime suggest the following possible links between them: (1) broken homes may be potentially criminogenic, but require 'catalytic agents' for their relationship to criminal behavior to develop; (2) certain types of broken homes may lead to crime; and (3) broken homes may be a consequence of conditions which tend to produce crime. The records of 230 men from 230 families living in factory-dominated communities in eastern Massachusetts who had been part of a counseling program designed to prevent delinquency were used in this study. Between 1939 and 1945, counselors recorded their observations of the sample, with subjects being between the ages of 9 and 16, and their families. Between 1975 and 1978, the subjects were traced (all but 3 percent were found), and court records were used to determine which had been convicted of serious crimes. The possibility that broken homes cause crime when 'catalytic agents' are present was tested by using parental conflict and aggressive models as possible catalytic agents. Data did not support this model. The possibility that broken homes have criminogenic effects when they lack certain socializing forces was tested by using absence of affection and absence of superision as relevant socializing forces. This model was not supported. Data did support the proposition that broken homes and crime are produced by similar conditions, but broken homes and crime are not causally related. Tabular data, 8 footnotes, and 13 references are provided.