Although earlier studies have indicated an inverse association between corpulence during childhood and the risk for breast cancer in adulthood, no study had evaluated body shape at different ages or body shape evolution in relation to the risk for breast cancer defined by menopausal status and hormone receptor status. Hazard ratios (HR) were estimated using multivariate Cox models in 81,089 women from the French E3N cohort to evaluate the risk for breast cancer associated with body shape (Sørensen's scale) at ages 8, menarche, 20-25, and 35-40 years. Six lifetime body shape trajectories were also estimated according to Nagin's approach to group-based trajectory modeling, and were analyzed in relation to the risk for breast cancer. During follow-up, 3573 breast cancer cases were diagnosed. At age 8 and at menarche, an increasing size of the body was associated with a significantly negative risk for estrogen receptor (ER)+/progesterone receptor (PR)+ postmenopausal breast cancer [both had a P(trend)=0.001; HR=0.80 (0.67-0.94) and HR=0.74 (0.64-0.86), respectively], for the largest body shapes compared with the first body shape. No significant association with body shape was found at the other ages. In terms of body shape evolution throughout life, women with a large body shape at menarche, irrespective of their body shape before or after menarche, tended to have a decreased risk for ER+/PR+ postmenopausal breast cancer compared with women who were constantly lean. Adjustment for current BMI or age at menarche did not modify any associations. Our study found an inverse association between adiposity at menarche and the risk for ER+/PR+ postmenopausal breast cancer, independent of corpulence at other ages. This work should encourage longitudinal studies with hormonal information to examine the underlying biological mechanisms.