The sample included children of schizophrenic, affectively ill, and normal parents followed since the ages of 7 to 12. Of 208 children, 159 were followed over time, and 36 were classified as dysfunctional. They were classified as dysfunctional because they had one or more hospitalizations for psychiatric reasons or because they received more than 6 months of treatment for serious psychiatric problems. Among children of schizophrenic parents, 45 percent were dysfunctional; percentages for children of parents with affective disorders and children of normal parents were 24 percent and 10 percent, respectively. Although subjects classified as dysfunctional in their mid-20's tended to show poor adjustment since early adolescence, individual subjects followed widely differing courses, indicating there are both straight and devious pathways to their current status. For subjects with erratic courses, their status as dysfunctional was very much related to when the classification was made. The authors conclude that risk research may require longer followup to evaluate stressors and buffers, along with genetic influences, that shape an individual's evolving outcome. 24 references, 5 tables, and 3 figures.