Although type 2 diabetes contributes to the risk of development of certain tumors; factors which modify this relationship need to be further examined. Familial diabetes frequency was followed in (a) diabetes-free cancer patients (132), aged 60.3 +/- 1.2; (b) cancer patients with apparent (62.3 +/- 0.4; n=371) or occult type 2 diabetes (61.6 +/- 1.2; n=65) and (c) cancer-free patients with type 2 diabetes (61.8 +/- 0.4; n=237); the latter group (172 females and 65 males) featured almost identical diabetes frequencies: 30.8 +/- 3.5% and 30.8 +/- 5.7%, respectively. They significantly exceeded a relevant index occurring in families free from cancer concomitant with diabetes (6.1 +/- 2.1%). Significantly lower frequencies of familial diabetes were reported in cancer patients as compared with cancer-free patients (12.9 +/- 2.1% females, p < or = 0.01; 19.4 +/- 3.8% - males, p=0.1). To sum up, it cannot be ruled out that familial diabetes is a marker of relative decrease in potential risk of cancer rather than that of increase. The causes of this phenomenon and its possible dependence on gender need to be explored.