Dyadic interactions between 84 at-risk toddlers and their nondrug-, polydrug-non-cocaine-, or polydrug-cocaine-using mothers were examined during unstructured play. Coded videotaped behaviors revealed two maternal constructs (interactive competence, maladaptive engagement) and three toddler constructs (responsiveness, active positive initiation, play). Cocaine-using mothers displayed more
... [Show full abstract] maladaptive engagement than did prenatal nondrug users; however, poorer interactive competence further distinguished them from polydrug-but-noncocaine users. Toddlers of cocaine-using mothers displayed less responsiveness and positive initiation than non-drug-group peers; however, these behaviors did not reliably differ from polydrug-non-cocaine counterparts. Play interest remained relatively similar across groups. Results suggest that difficulties in social interaction appear to continue among cocaine-using mothers during dyadic exchanges with their offspring in the second year; however, toddler deficits do not appear to be greater than those experienced by polydrug-non-cocaine counterparts.