Clinical and genetic studies of familial nonmedullary thyroid cancer (FNMTC) have yielded conflicting results concerning the aggressiveness of the tumors, and uncertainty of their genetic makeup. In most reports of multiply affected families, the composition of the kindreds has favored families of 2 affected members. Using data for differentiated thyroid cancer (DTC) provided by the Surveillance Epidemiology and End Results (SEER) branch of the National Cancer Institute, and fine-needle aspiration data from Mayo Clinic, I found that the likelihood of 2 cases of sporadic DTC (RR) in a 9-member first-degree family was 1.25% of all DTC families, amounting to 39.4% of 306 multi-hit families reported in the literature. To study the remaining affected families I used the Bernouilli trials model of exact probability. The 60.6% of non-RR, multiply affected families are mostly concentrated in kindreds of 2 to 5 affected members. In 2-hit families, 62%-69% of affected members are sporadic (RR) cases. In families having 3 or more affected members, fewer than 6% have 1 or more sporadic (R) cases, and fewer than 0.15% have 2 or more. In families of 3 to 5 affected members, more than 96% of affected members have the familial (F) trait. Approximately 1 of 338 DTC cases carries the F-trait. Since approximately 40% of multiply affected member first-degree kindreds of DTC, and a significant majority of 2-hit families, are composed of clinically evident, sporadic cases only clinical and genetic investigations of FNMTC should center on families of 3 or more affected members.