We analysed the effects of number of siblings on the risk of solid tumours using the Swedish Family-Cancer Database, including population-based information on over 11 million individuals and more than 178 000 cancer patients diagnosed between 1958 and 2004. Incidence rate ratios (RRs), estimated by Poisson regression models, were adjusted for age, sex, birth cohort, area of residence and socioeconomic status. Having eight or more siblings vs none increased the risk of stomach cancer (RR ¼ 1.83, 95% confidence interval (CI), 1.44 -2.34). Anal cancer diagnosed before age 40 showed the strongest association with the total siblings (RR ¼ 3.27, 95% CI, 2.04 -5.26 for five or more siblings vs none). Endometrial (RR ¼ 0.76, 95% CI, 0.70 -0.82), testicular (RR ¼ 0.71, 95% CI, 0.62 -0.82), skin cancer (RR ¼ 0.82, 95% CI, 0.69 -0.97) and melanoma (RR ¼ 0.72, 95% CI, 0.65 -0.79) showed strong decreased risks for five or more siblings vs none. Prostate cancer risk for those with five or more older siblings vs none was 1.38 (95% CI, 1.23 -1.55). Having five or more younger siblings was most strongly associated with stomach cancer (RR ¼ 1.59, 95% CI, 1.29 -1.95) and melanoma (RR ¼ 0.68, 95% CI, 0.59 -0.79). We conclude that sibship characteristics are strong correlates of cancer risk at several sites; plausible interpretations include socioeconomic status.