Kentucky's PACE program, which was designed to provide adult basic education and parenting skills training for parents, early childhood education for children, and parent-child interaction and learning, lacks proof for the validity of its theoretical base and service model. Questions that need to be resolved before widespread replication of the PACE program occurs include the following: (1) Is there a causal relationship between family background and children's educational and social outcomes? (2) Will changes in family attitudes and behaviors change children's outcomes? (3) Can replication of the program in different settings by different providers reproduce original effects? Program review produced a statement of the theoretical model implied by the program's assumptions. The model consisted of hypotheses about family background, children's outcomes, and PACE services. Drawing from the literature on family background and child learning, an explanatory model includes hypotheses about the causes of undereducation. A prevention and remediation model includes hypotheses about the effect of PACE services on the amelioration of literacy problems related to family factors. The present paper presents these hypotheses along with ideas about how they might be tested. Concluding material offers a strategy for evaluating the program. (RH)