The frequency of vaginal thrush was determined in an unselected group of patients attending the antenatal clinic. Diagnosis was based on phase-contrast microscopy and culture. One hundred and two patients were initially included but only 98 completed this study. Thrush infection was detected in 25 (25%) patients. All 25 patients underwent treatment with one vaginal ovule containing 400 mg ketoconazole on each of three successive evenings. Thrush attack was eliminated in 23 (92%) cases. In view of the virtual absence of clinical symptoms, the recommendation to undertake general screening for thrush in pregnancy is discussed. The efficacy of treatment with ketoconazole ovules was studied and the results of treatment were reviewed after 4 and 8 weeks. Recurrence or re-infection has to be considered in pregnancy so that thrush treatment should sensibly be undertaken in the third trimester unless subjective symptoms indicate the need for earlier therapy.