OBJECTIVE--To determine the effect of atenolol on the outcome of pregnancy in women with essential hypertension. DESIGN--Prospective, randomised, double blind, placebo controlled study. SETTING--Hospital clinic. PATIENTS--33 Women with mild essential hypertension (systolic blood pressure 140-170 mm Hg or diastolic pressure 90-110 mm Hg on two occasions at least 24 hours apart) consecutively referred to two obstetric medical clinics. Four patients in the placebo group were withdrawn from the study: control of blood pressure was inadequate in two, one developed breathlessness, and one changed her mind about participating. The mean gestation in the 29 remaining women on entry to the study was 15.9 weeks. MAIN OUTCOME MEASURES--Blood pressure and birth weight. INTERVENTION--14 Women received placebo. 15 Women received atenolol 50 mg daily initially, increasing until either the blood pressure was less than 140/90 mm Hg or a dose of 200 micrograms daily was reached. RESULTS--The mean blood pressure on entry was 148/86 mm Hg in the group given atenolol and 144/86 mm Hg in the group given placebo. During treatment the mean diastolic pressure was significantly reduced by atenolol compared with placebo (to 74 v 81 mm Hg; difference in means (95% confidence interval) 7.0 (2.9 to 10.0) mm Hg) but the effect on systolic pressure was marginal (132 v 136 mm Hg; 4.0 (-1.4 to 8.6) mm Hg). Babies in the atenolol group had a significantly lower birth weight than those in the placebo group (2620 g v 3530 g; 910 (440 to 1380)g). CONCLUSION--Atenolol given from the end of the first trimester in patients with mild hypertension is associated with intrauterine growth retardation. When taken in conjunction with the results of a previous study in which methyldopa was given these findings indicate that benefit is unlikely to result from treating mild essential hypertension in pregnancy.