In a single practice during the 21 years 1971-1991, the incidence of gestational diabetes in pregnancies in which norethisterone was prescribed was 32.4% (22 of 69) in comparison with 7.1% in pregnancies in which the women did not take norethisterone (137 of 1,684) (p < 0.001). Gestational diabetes was no less severe (degree of hyperglycaemia, need for insulin therapy) when associated with norethisterone. However, follow-up revealed that gestational diabetes when associated with norethisterone had a lesser risk of emerging diabetes mellitus and impaired glucose tolerance. Masculinization of a female fetus occurred in 5 of 39 (12.8%) exposed to norethisterone; all were cases of clitoral hypertrophy not requiring surgical treatment. Norethisterone in these 69 pregnancies accounted for 33.3% (5 of 15) cases of clitoral hypertrophy diagnosed in 100,756 consecutive births.