The influence of ethnic origin, body mass index, and parity on the frequency of gestational diabetes was assessed in 11,205 consecutive women attending a multiracial antenatal clinic in London, where all women were screened for gestational diabetes. Logistic regression was used to model the relationship between gestational diabetes and ethnic origin, age, body mass index (BMI), and parity. Results were presented as adjusted odds ratios, where the reference categories are White women, age < 25 years, BMI < 27, and parity < 3. Ethnic origin was the dominant influence on the prevalence of gestational diabetes. Women from ethnic groups other than White had a higher frequency of gestational diabetes than White women (2.9% vs 0.4%, p < 0.001). Compared to White women the relative risk of gestational diabetes in the other ethnic groups was: Black 3.1 (95% confidence limits 1.8-5.5), South East Asian 7.6 (4.1-14.1), Indian 11.3 (6.8-18.8), and miscellaneous 5.9 (3.5-9.9). Increasing age was an independent risk factor. The relative risk was higher in women > or = 35 years in all ethnic groups other than in South East Asian women. Obesity (BMI > or = 27) was a further independent risk factor in all ethnic groups except in the Indian and South East Asian women. Parity > or = 3 increased the relative risk of gestational diabetes in the White, Black, and South East Asian women only.(ABSTRACT TRUNCATED AT 250 WORDS)