Abstract The effect of a birth companion on duration of labor was examined in a sample of 66 nulliparous women in Michigan. Two-thirds of the sample were single, 66% received Medicaid, and 88% were white. Using simple linear regression, support during labor explained a significant amount of variance in duration of labor. Duration of labor for 8 unsupported mothers was significantly longer (965 ± 206 min) than that for 58 supported mothers. Using hierarchic regression and including maternal education, marital status, race, amniotomy, and labor induction in the model before support, support during labor continued to explain a significant amount of variance in duration of labor. These findings are similar to those reported from earlier randomized trials of social support during labor. Evidence continues to accumulate that confirms the proposition that labor duration is affected by environmental as well as biologic factors.