The present randomized controlled trial tested the effectiveness of a behavioral counseling program for reducing children's exposure to environmental tobacco smoke (ETS). Counseling was delivered by clinic staff as part of well-child health care services in a community clinic setting. A total of 150 mothers with children aged 4 years or younger were recruited. Parent-reported and children's urinary cotinine measures of ETS exposure were obtained at baseline, 3 months, 6 months (post-test), and 12 months (follow-up). Saliva samples were obtained from mothers who reported quitting smoking, for objective verification by thiocyanate analysis. After baseline, mothers were randomly assigned to a measures-only control condition or an intervention consisting of seven behavioral counseling sessions over 6 months. Counseling included behavioral contracting, self-monitoring, problem solving, and positive reinforcement. Results indicated acceptable test-retest reliability and validity of measures. Parent-reported measures indicated that, in both groups, children's exposure to their mothers' tobacco smoke in the home and to all tobacco smoke declined steeply from baseline to 6 months post-test, and remained essentially level during follow-up. Mothers' smoking rates followed the same pattern. Children's urinary cotinine concentrations did not show significant change over time in either group. Findings on the fidelity of treatment implementation suggest that the structure and funding of the community clinic health care system and associated staff turnover and training issues resulted in participants receiving a less efficacious intervention than in our past efficacy trials. Implications for future effectiveness trials are discussed.