Seventy-one college general biology students were taught a unit in Mendelian genetics by the traditional lecture method. Emphasis was placed on meiotic formation of gametes, the Law of Segregation, and the Law of Independent Assortment. The Punnett-square model was used for all practice problems. Eight weeks later, a content-validated retention test was given to evaluate the students' retention of problem-solving skills. The test required students to use proportional reasoning (identifying ratios from the Punnett squares), combinatorial reasoning (identifying combinations of gametes from parental genotypes), and probabilistic reasoning (estimating gamete or offspring probabilities). Each of the 71 students was also given three Piagetian interview tasks to evaluate intellectual development in the areas of reasoning under question. The balance-beam task, the electronic switch-box task, and colored squares and diamonds were used to test for proportional reasoning, combinatorial reasoning, and probabilistic reasoning, respectively. Pearson correlations and factor analysis failed to show direct relationships among Piagetian tasks for the three kinds of reasoning and their corresponding occurrence in genetics problems. Some correlations were higher between different reasoning types than between similar types. Analysis of variance showed significant differences for all three reasoning types among concrete-operational, transitional, and formal-operational students with the retention test. Post-hoc analysis of ANOVAs indicated that formal-operational students had significantly more success in the three reasoning areas than transitional students, and transitional students had significantly more success than concrete-operational students.