Descriptions of the problem-solving strategies of experts solving realistic, computer-generated transmission genetics problems are presented in this paper and implications for instruction are discussed. Seven experts were involved in the study. All of the experts had a doctoral degree and experience in both teaching and doing research in genetics. Two types of data were available for analysis and for the description of the strategic knowledge that was used by the experts. These were the transcripts of the think aloud protocols and the computer printouts of the sequence of crosses for each genetics problem. Tables are provided which summarize the experts' strategies. Implications for instruction in solving genetics problems are reviewed in the areas of: (1) the utility of the model used for the study of problem solving; (2) the content knowledge of expert problem solvers; (3) clear and explicit information on use of computer-generated problems; and (4) what strategic knowledge to teach and when and how to teach it. A flowchart of the solution path used by experts to solve genetics problems is also included. (ML)