In the present paper we compared the results of the application of segregation analysis, under two different single major locus (SML) transmission hypotheses, a dominant one with sex effect and a recessive one, to the families of 202 probands with major depression, recurrent and bipolar disorder. In the first analysis we considered only secondary cases with major affective disorders (bipolar disorders and major depression, recurrent), in the second one we included as affected phenotypes also relatives with atypical depression, dysthymic and cyclothymic disorders. Results indicated that considering spectrum disorders greatly modified familial segregation patterns.