Individual-to-group (IG), group-to-individual (GI), individual-to-individual (II), and group-to-group (GG) transfer were compared in two experiments. In Experiment 1, groups had fewer instances to solution than individuals on training problems, and the pattern of instances on the transfer problems was (IG = GG) < (GI = II). Experiment 2 replicated both findings with more difficult problems. Thus, previous individual experience did not influence subsequent group performance, and previous effective performance in groups did not transfer to subsequent individual performance. Rather, groups performed better than individuals regardless of the temporal order of individual and group problem-solving experience.The effect of previous individual problemsolving experience upon subsequent group problem solving and of previous group problem-solving experience upon subsequent individual problem solving is a fundamental question in both experimental psychology and educational and social policy. However, to our knowledge, surprisingly few studies have addressed either of these questions. Two experiments (Dunnette, Campbell, & Jaastad, 1963;Perlmutter & de Montmollin, 1952) compared the temporal order of individual and group experience for the same subjects working both alone and in groups, concluding that the group first-individual second (GI) order resulted in better individual performance than the individual firstgroup second (IG) order. However, their conclusion that previous group problemsolving experience facilitates subsequent individual problem solving is only suggestive because neither study included either individual-to-individual (II) or group-to-