Lead-contaminated house dust is one factor in childhood lead poisoning; however, most lead-reduction programs do not emphasize the control of house dust. We studied whether lead-reduction plus dust-control measures would lower blood lead levels in children with Class II or III poisoning (blood lead levels, 30 to 49 micrograms per deciliter) more effectively than lead reduction alone. An experimental group of 14 children and a control group of 35 children whose homes had already been treated were studied. In experimental homes, sites with elevated lead levels (greater than 100 micrograms per 930 cm2) were wet-mopped twice monthly and families were encouraged to clean and to wash the child's hands frequently. After one year blood lead levels fell an average of 6.9 micrograms per deciliter in the experimental group, as compared with 0.7 micrograms per deciliter in controls (P less than 0.001). Children in the experimental group with the highest blood lead levels had the most marked reduction. Four children in the control group (and none in the experimental group) required chelation therapy for blood levels greater than 50 micrograms per deciliter. These results show that a focused dust-control program can reduce blood lead levels more than standard lead removal in the home.