We examined the utility of naive ratings of communication patterns and relationship quality in a large sample of distressed couples. Untrained raters assessed 10-min videotaped interactions from 134 distressed couples who participated in both problem-solving and social support discussions at each of 3 time points (pre-therapy, post-therapy, and 2-year follow-up) during a randomized clinical trial of behavioral couple therapy. Teams of naive raters observed a particular type of discussion from the 3 time points at a single sitting in a random order and rated dyadic interaction patterns (negative reciprocity, positive reciprocity, wife demand/husband withdraw, husband demand/wife withdraw, and mutual avoidance) and the overall relationship quality of couples. These naive ratings were strongly and consistently associated with both levels of, and changes in, trained observational codes and self-reported relationship satisfaction. Naive ratings of couples accounted for similar--and at times superior--amounts of variance in both concurrent relationship satisfaction and divorce at 5-year follow-up when compared with trained ratings. These findings offer compelling support for the use of naive raters in research with couples and also suggest important future directions that are applicable to both research and practice with distressed couples. (Contains 9 tables and 4 footnotes.)