This study investigated the efficacy of Early Head Start home-based, center-based and mixed-approach programs on cognitive, language and behavioral outcomes at different levels of cumulative environmental risk. Early Head Start is a federal program that provides low-income families and their children from birth to age three with childcare, parenting education, healthcare and other family supports. As part of Early Head Start's initiation, a program evaluation was begun involving 3,001 children from 17 programs around the country. Half of the children were randomly assigned to the control group, who received no Early Head Start services. Data were collected through program application and enrollment forms, interviews of parents and child and family assessments. Almost all of the children's primary caretakers were mothers, ranging in age from 18 to 26. One-third were African American, one-third white, and one-fourth Hispanic. Almost half of the parents did not have a high school diploma at the time of enrollment, and most of the families received public support of some kind. For each child, a multiple environmental risk score was calculated, which was the sum of 10 possible environmental risks. Each of four outcomes was regressed onto the ten risks individually and also as a cumulative risk index along with program type and covariates. There were significant negative relations of accumulated risk to reductions in reasoning, spatial ability and vocabulary and increased behavior problems. Children with at least eight risks scored 1.48 standard deviations lower on reasoning ability and vocabulary, .48 standard deviations lower on spatial ability and .48 standard deviations higher on behavior problems. The home-based program showed significant benefit for reasoning and vocabulary. Versus the control group, home-based programs increased average reasoning scores by .24 of a standard deviation and increased vocabulary by .14 of a standard deviation. There was no significant difference in program benefits at different levels of risk. This suggests that for reasoning and vocabulary, the home-based program is promotive because the degree of benefit Early Head Start appears to provide is consistent across all levels of risk for the set of risks and outcomes examined in this study.