Hypotheses presented in an initial report of this study were tested and, in general, all were supported. Maternal diagnosis, chronicity, and age of onset with regard to the child and to parent/child interaction were significantly associated with child competence. However, effects of these factors were not consistent across age of child or child environment. Affective diagnosis and chronicity of emotional illness were found to have strong associations with competence. When school competence ratings were used to define high versus low competence, affective diagnoses of mothers hospitalized for psychiatric disorders appeared to have a great effect on older children (10 years old), while chronicity of mothers' emotional illness more highly affected younger children (7 years old). When parent ratings were used to define high versus low functioning, affective diagnoses of mothers were significantly related to competence of children at both ages, while chronicity of illness again was related to competence only in the younger group. Indexes of balance and warmth of parent/child interactions were significantly related to child competence when defined by school ratings, but not when defined by parent ratings. Data are discussed in relationship to the need for multiple indexes of competence and with respect to the requirements of children at risk in different age groups. Also, suggestions are offered regarding the compensatory effect of the nondisturbed parent through the influence of an affective family interaction construct. (Author/RH)