Tested a structural model (Patterson, 1986) of the effects of parental discipline on child coercive and antisocial behavior using a sample of 65 recently separated single-mother families with 6- to 8-year-old boys. Previous tests of this model used older boys and samples that allowed family structure to vary. The replication model produced acceptable fit, chi-sup-2(61, N = 65) = 70.6, p = .188, and explained 11% of the variance in antisocial behavior. Five alternative models were tested on this sample and are discussed briefly. The significance of generalization to this sample is discussed for the model’s clinical usefulness, and further improvements to increase explainable variance are suggested.