Media consumption may contribute to childhood obesity. This study developed and evaluated a theory-based, parent-focused intervention to reduce television and other media consumption to prevent and reduce childhood obesity. Families (n = 202) with children ages 6-9 were recruited from a large, urban multiethnic population into a randomized controlled trial (101 families into the intervention group and 101 into the control group), and were followed for 6 months. The intervention consisted of a 2-hour workshop and six bimonthly newsletters. Behavioral objectives included: (i) reduce TV watching; (ii) turn off TV when nobody is watching; (iii) no TV with meals; (iv) no TV in the child's bedroom; and (v) engage in fun non-media related activities. Parents were 89% female, 44% white, 28% African American, 17% Latino, and 11% Asian, mean age 40 years (s.d. = 7.5); 72% were married. Children were 49% female, mean age 8 years (s.d. = 0.95). Sixty-five percent of households had three or more TVs and video game players; 37% had at least one handheld video game, and 53% had three or more computers. Average children's weekday media exposure was 6.1 hours. At 6 months follow-up, the intervention group was less likely to report the TV being on when nobody was watching (adjusted odds ratio (AOR) = 0.23, P < 0.05), less likely to report eating snacks while watching TV (AOR = 0.47, P < 0.05), and less likely to have a TV in the child's bedroom (AOR = 0.23, P < 0.01). There was a trend toward reducing actual media consumption but these outcomes did not reach statistical significance. Effective strategies to reduce children's TV viewing were identified.