Objective To determine the nature and outcome of obstetric cholestasis in a United Kingdom population. Design Prospective analysis of clinical outcome in women diagnosed with obstetric cholestasis that is actively managed. Setting Antenatal population of three London hospitals between August 1999 and April 2001. Population Seventy women with obstetric cholestasis defined as abnormal liver function (one or more abnormality in gamma-glutamyl transpeptidase, alanine amino-transferase, aspartate amino-transferase and total bile acids) in a pregnant woman with pruritus, in the absence of other pathology. Methods All women were interviewed weekly regarding their symptoms. All were actively managed according to a standardised protocol, which included early delivery before 38 weeks. Obstetric outcome was recorded. Results Seventy women of mean age 30 (6) years delivered 73 infants. The median gestation at onset of pruritus was 30 (range 4-39) weeks and at diagnosis of obstetric cholestasis was 33.7 (range 21-40.7) weeks. Asian women were more likely to be diagnosed with obstetric cholestasis. Pruritus was usually severe and generalised, and commonly worst on the palms and/or soles of the feet. There were no stillbirths or perinatal deaths. Twenty-five women required caesarean section (36%); only four (16%) were for fetal distress. Twelve women (17%) delivered before 37 weeks, of which eight (67%) were iatrogenic. Ten (14%) infants required admission to the special care baby unit of which four (40%) were ventilated. Conclusions Policies of active management result in increased intervention and associated complications. This must be balanced against possible reductions in perinatal mortality.