Using the Swedish Familial Cancer Database, organized and studied by the Hemminki group since 1997, the hypotheses that breast, prostate and colorectal cancers were affected by familial risks in Sweden (1958-2008) are tested. The null hypotheses were rejected for the three cancer types up to the age interval 70-74 yrs for which probative data are available. Familial risks, represented as age-specific mortality rate ratios, are found to be age invariant: 1.9±0.3 (95% CI) for sons or daughters ages 45 to 74 of parents dying of colorectal cancer, 2.0 ± 0.2 for sons ages 50 to 74 of fathers dying of prostate cancer. The ratio of the risks for daughters of mothers dying of breast cancer ages 35 to 74, suggests a different behavior between pre-menopause and post-menopause ages. Can be seen as constant at 2.3± 0.3 for all observed age groups. The data could also be interpreted as showing a different behavior between the ages before and after menopause. The ratio could be decreasing linearly before menopause and be constant after menopause. One potential risk parameter is eliminated as familial because this parameter is identical for the general and familial populations: growth rate of preneoplastic lesions. Initiation or promotion oncomutation rates in stem cells could account for the familial risks in accord with wide variations of mutation rates observed in developing colon and lungs. Population parameters such as the fraction of persons in whom initiated stem cells would continue to grow after maturity could also have caused the observed familial risks, with oncomutation rates. For familial colorectal cancer, the two population parameters considered are of little effect, permitting the conclusion that familial colorectal cancer risk lies solely in oncomutation rates. Studies of immigrants have demonstrated that fetal/juvenile environmental factors are major determinants of organ-specific adult cancer risk. From these considerations emerges the hypothesis that the prime determinants of colorectal and other cancer risks lie in the fetal/juvenile oncomutation rates. This hypothesis may be tested directly by measuring and comparing somatic mutations at defined genetic loci (a.) in parents and their adult children (b.) among all adults with and without cancers in a specified organ.