The authors empirically examined the General Theory of Crime by Gottfredson and Hirschi (1990), with revisions suggested by the work of Sampson and Laub (1993). The generalized form of risk-taking behavior examined in the current study included elements of delinquency, including theft and assault, as well as tobacco, alcohol, cannabis, and hard drug use. Data were obtained from a single stratified sample of 1,075 students in Canada in grades 7, 9, 11, and 13. The findings are generally consistent with Sampson and Laub's approach to the General Theory of Crime. Family structure has an indirect effect on risk-taking. Family attachment has a direct effect on risk-taking, but risk-taking is also independently associated with peers (positively) and school attachment (negatively). The most interesting finding is that the effect of family attachment on risk-taking is moderated by both school and peer involvement. When family attachment is low, school attachment inhibits risk-taking, and strong peer attachment reinforces it. This suggests a role for social capital beyond the family -- both positive and negative -- in the production of risky behavior during adolescence. 2 tables, 4 figures, and 34 references