The relationship between age at birth of a first child and breast cancer was evaluated for 1159 affected women and 11,590 women without cancer in data collected in 1976 among married female registered nurses residing in 11 states in the United States. A positive trend of increasing risk of breast cancer with later ages at first birth was found (chi 2(1) for trend in proportions = 30.9, p less than 0.01). Adjustment for potential confounding variables by multiple logistic regression did not affect this trend. The presence of this relationship using non-hospitalized controls of similar social status to cases supports the reality of this association, which has recently been challenged as an artifact due to inappropriate choice of hospitalized controls.