Developmental evaluations of the current wave of welfare reform programs present challenges with regard to (1) assessing child outcomes; (2) accounting for heterogeneity among low-income families in both baseline characteristics and involvement in self-sufficiency activities and supports, and (3) development of alternatives to experimental approaches to causal inference. This study (N = 1,079) addresses these challenges by examining effects on 4- to 6-year-old children of different patterns of child care, self-sufficiency activities, and other service utilization indicators among experimental-group mothers in a 16-site welfare reform program. Outcomes in areas of cognitive ability and behavior problems were investigated. The study identified seven subgroups of participants engaging in different patterns of service utilization and activity involvement. A two-stage simultaneous equation methodology was used to account for selection, and effects on child cognitive ability of participation in specific patterns of services and activities were found. For example, children of mothers characterized by high levels of involvement in center-based child care, education, and job training showed higher levels of cognitive ability than children of mothers in groups characterized by high involvement in center-based care and education, or center-based care and job training. In addition, children of mothers in groups with high levels of involvement in any of these activities showed higher levels of cognitive ability than those with low levels of involvement. The bulk of selection effects occurred through site-level differences, rather than family-level socio-economic status or maternal depression indicators. Implications for welfare reform program and policy concerns are discussed.