Objective: The aim of the study was to characterise the incidence, type and presentation of tuberculosis in pregnancy over a 5-year period in women booked for antenatal care in a District General Hospital located in a high prevalence area in London. We also aimed to identify any problems and difficulties in the diagnosis and management of tuberculosis associated with pregnancy. Design: Retrospective review of computer records and hospital notes over a period of 5 years from January 1997 to December 2001. Demographic and clinical data were collected for all the cases identified. Population: All women with tuberculosis who conceived on antituberculous treatment, or had onset of symptoms or diagnosis made in pregnancy or the immediate postpartum period(6 weeks), and booked for antenatal care at a District General Hospital located in an area of high prevalence for tuberculosis(52.2 per 100,000 population in Ealing, Hammersmith and Hounslow Health authority according to the National Tuberculosis Survey of England and Wales in 1998). Results: Thirty-two women were identified over the 5-year period, giving an incidence of 252/100,000 deliveries. The number of cases increased from 3 in 1997 to 10 in each of 2000 and 2001. All of these women were from ethnic minorities and 88%of them were immigrants with the median interval from arrival in UK being 2 years. Fifty-three percent were diagnosed with extrapulmonary tuberculosis, 38%with pulmonary tuberculosis and 9%had both. The median duration of symptoms prior to presentation was 31 days(being longer in women with extrapulmonary tuberculosis); the longest was 10 years. The median interval from presentation of symptoms to diagnosis was 32 days and the majority of women started treatment immediately. The commonest reason for a delay in diagnosis was late presentation(52%),followed by non- specific symptoms(in 38%). There was a trend towards late presentation among recent immigrants(odds ratio 2.14, 95%confidence interval 0.44-10.53) and those having extrapulmonary tuberculosis(odds ratio 1.64, 95%CI 0.32-8.45). Most of the women(28/32) showed good compliance and a good response to treatment(31/32). Maternal outcomes were good with no serious morbidity or mortality. The majority of women delivered at term(28/32), while two delivered preterm and two women miscarried. There was no perinatal mortality. Conclusion: This is the largest recent series of pregnant women with tuberculosis in the UK. A high incidence of extrapulmonary tuberculosis was seen. The common causes for a delay in diagnosis were late presentation and non-specific symptoms. If recent immigrants from high prevalence areas who have been in the UK for less than 5 years were asked for symptoms suggestive of tuberculosis at the booking visit and through pregnancy, this might facilitate early diagnosis and treatment. The use of a symptom questionnaire at the booking visit for these women could be a method to alert both the women and health professionals involved in their care, to the symptoms of tuberculosis. With good compliance, there is a good response to treatment and favourable maternal and perinatal outcomes. Objective: The aim of the study was to characterise the incidence, type and presentation of tuberculosis in pregnancy over a 5-year period in women booked for antenatal care in a District General Hospital located in a high prevalence area in London. We also aimed to identify any problems and difficulties in the diagnosis and management of tuberculosis associated with pregnancy. Design: Retrospective review of computer records and hospital notes over a period of 5 years from January 1997 to December 2001. Demographic and clinical data were collected for all the cases identified. Population: All women with tuberculosis who conceived on antituberculous treatment, or had onset of symptoms or diagnosis made in pregnancy or the immediate postpartum period (6 weeks), and booked for antenatal care at a District General Hospital located in an area of ​​high prevalence for tuberculosis (52.2 per 100,000 population in Ealing, Hammersmith and Hounslow Health authority according to the National Tubercu losis Survey of England and Wales in 1998). Results: Thirty-two women were identified over the 5-year period, giving an incidence of 252 / 100,000 deliveries. The number of cases increased from 3 in 1997 to 10 in each of 2000 and All of these women were from ethnic minorities and 88% of them were immigrants with the median interval from arrival in UK being 2 years. Fifty-three percent were diagnosed with extrapulmonary tuberculosis, 38% with pulmonary tuberculosis and 9% had both. The median duration of symptoms prior to presentation was 31 days (being longer in women with extrapulmonary tuberculosis); the longest was 10 years. The median interval from presentation of symptoms to proof was 32 days and the majority of women started treatment immediately. The commonest reason for a delay in diagnosis was late presentation (52%), followed by non-specific symptoms (in 38%). There was a trend towards late presentation in recent immigrants (odds ratio 2.14, 95% confidence interval 0.44-10.53) and those with extra pulmonary tuberculosis (odds ratio 1.64, 95% CI 0.32-8.45). Most of the women (28/32) showed good compliance and a good response to treatment (31/32). Maternal outcomes were good with no serious morbidity or mortality. The majority of women delivered at term (28/32), while two delivered preterm and two women miscarried. There was no perinatal mortality. Conclusion: This is the largest recent series of pregnant women with tuberculosis in the UK The common causes for a delay in diagnosis were extradial tuberculosis was seen. The common causes for a delay in diagnosis were late presentation and non-specific symptoms. of tuberculosis at the booking visit and through pregnancy, this may facilitate early diagnosis and treatment. The use of a symptom questionnaire at the booking visit for these women could be a method to alert both the women and health professiona ls involved in their care, to the symptoms of tuberculosis. With good compliance, there is a good response to treatment and favourable maternal and perinatal outcomes.