In this study, we examined when and whether married people engage in attributional activity or form causal attributions to explain their partners' behavior. We used an indirect probe to better approximate naturally occurring cognitive activity. We also examined the content of spouses' causal attributions, using both direct and indirect probes. Spouses were asked about frequent as well as infrequent relationship events, and about partner behaviors that had positive or negative impacts on the recipient. Husbands in unsatisfying relationships reported more attributional thoughts than did happily married husbands, whereas wives in the two groups did not differ. Behaviors having negative impacts elicited more attributional activity than did positive behaviors. Behavioral frequency and impact interacted in ways contrary to predictions. Finally, distressed couples were particularly likely to report distress-maintaining attributions and were particularly unlikely to report relationship-enhancing attributions, compared with their nondistressed counterparts.