In a representative sample of the German population (N=2047), the first three dimensions of the Toronto Alexithymia Scale TAS-26, as well as a fourth scale of general alexithymia (scales 1 to 3, 18 items) were investigated in order to determine the prevalence and possible sociodemographic predictors of the construct. Participants were recruited by chance (random route technique) in their households and interviewed face to face. Sociodemographic predictors included v. o. sex, age, partnership, residency in East or West Germany, religious affiliation, workstatus and socioeconomic status (based on education, profession, income). A prevalence of 10.2% for alexithymia could be determined in this sample. Alexithymia in Germany is predicted by residency in East-Germany, lacking partnership, religious affiliation, work loss as well as by low socioeconomic status, which proved to be the strongest predictor. The multiple regression models accounted for 4 to 7% of the total variance (a-adj values). Alexithymia, in terms of variance explained, is weakly predicted by sociodemographic factors. This influence, however, proved to be highly significant. The construct, represented by the TAS-26, appears to be largely independent from age and sex. The difference between German residents in east and west in scale 3 of the TAS-26 may be associated with higher pressure towards social conformism in East- vs. West-Germany.