This study sought to construct a set of family strengths using data from the National Longitudinal Survey of Youth - Child Supplement (NLSY-CS), and, if the family strengths' construct was possible, to discern whether the presence of these family strengths predict better outcomes for children. The original NLSY sampled approximately 12,000 children, ages 14 to 21, in 1979; subjects have been surveyed every year since. The current analysis uses data from 1986 and 1988 to predict 1990 child behavior outcomes. The 1990 data come from 2 groups of children, the first including 277 six- to nine-year-olds, and the second, involving 1,163 ten- to fourteen-year-olds. Findings indicate that in both age samples there is considerable variability in the influence of family strength measures, and no significant difference in results across gender, racial/ethnic, or structural family lines. While it appears that children in families with higher family strength levels have fewer behavioral problems, the degree of statistical significance is modest among the younger children and lower among the older ones. Results, however, do point to a positive correlation between family strengths and fewer behavioral problems, especially when mother-partner measures are involved. Results suggest that additional studies are merited. (Two appendices list child outcomes and family well-being indicators; 28 tables graph family strength data. Contains 13 references.) (JC)