Taken together, the studies of Grasmick et al. and Keane et al. reinforce the value of exploring further the concept that people are differentially controlled by the long-term consequences of their acts. Their thoughtful efforts to understand and test the implications of the theory lead to consideration of several general issues: measuring self- control, sampling, self-control as propensity, and causal order. Multiple measures of self-control as used by Grasmick et al. are desirable, yet behavioral measures of self-control as employed by Keane et al. emerge as preferable to self- report. Similarly, the sampling approach adopted by Keane et al. is more consistent with testing requirements of the theory. Another advantage of behavioral measures is that they counter the tendency to translate the control concept at the core of the Hirschi and Gottfredson control theory of crime into a personality concept or "an enduring criminal predisposition." This aspect of the Grasmick et al. study proves the most disappointing tendency in response to the Hirschi and Gottfredson theory. An enduring predisposition to consider the long-term consequences of one's acts may be evident in the theory, but there is no personality trait predisposing individuals toward crime. In regard to causal order, Hirschi and Gottfredson support the stance of Keane et al. who suggest that causal order problems may be solved by conceptual analysis and by resort to collateral data. 1 footnote and 16 references