OBJECTIVE To investigate the observation that children with pervasive developmental disorders have later and more prolonged lead exposure and are more likely to be reexposed when compared to lead-poisoned children without pervasive developmental disorders. METHODS Retrospective chart review. METHODS A large, urban lead treatment program. RESULTS Over a six year period 17 children with pervasive developmental disorders (including autism) were treated. Compared to a randomly selected group of 30 children without pervasive developmental disorders who were treated for plumbism over the sam interval, those with pervasive developmental delay were significantly older at diagnosis (46.5 vs 30.3 months, p = .03) and had a longer period of elevated blood lead levels (39.1 vs 14.1 months, p = .013) during management. Despite close monitoring, state-mandated environmental inspection and prompt lead hazard reduction or alternative housing, 75% of children with pervasive developmental disorders were reexposed to lead during medical management compared with 23% of children without pervasive developmental disorders (p = .001). CONCLUSIONS 1) lead intoxication among children with pervasive developmental disorders may appear de novo beyond the third year of life and is associated with a high rate of reexposure; 2) the provision of deleaded housing (by current techniques) may not be sufficient to protect these children from repeated lead exposure; 3) these data support recommendations by the Centers for Disease Control that children with developmental delays be closely monitored for the appearance of lead intoxication. This monitoring should continue beyond the third year of life.