A study was made to determine whether learning achieved in a T-group experience is transferred to a problem-solving situation, and to identify possible factors in the training process that enhance transfer of training. Problem-solving was measured along several dimensions, including quality of solution offered, degree of acceptance each member gave to a group solution, cohesiveness of groups, degree of status consensus within groups, perceived amounts of conflict within groups, and interaction rates. Experimental subjects were exposed to T-group training with or without theoretical lectures; control subjects were not exposed to T-group training. There was also a problem-solving session in which all groups participated. Data were gathered by means of a postsession questionnaire, interaction observations, and solution analysis. Results of analysis of variance indicated that T-group training does produce behavioral changes in trainees and that such changes can be transferred to another situation; but the view that cognitive learning enhances transfer of training was not supported. It is possible, however, that the lectures were not long enough to have a measurable effect. (Illustrations, appendixes, and 88 references are included.) (author/ly)