BackgroundBehavioural interventions might improve clinical outcomes in pregnant women who are obese. We aimed to investigate whether a complex intervention addressing diet and physical activity could reduce the incidence of gestational diabetes and large-for-gestational-age infants.MethodsThe UK Pregnancies Better Eating and Activity Trial (UPBEAT) is a randomised controlled trial done at antenatal clinics in eight hospitals in multi-ethnic, inner-city locations in the UK. We recruited pregnant women (15–18 weeks plus 6 days of gestation) older than 16 years who were obese (BMI ≥30 kg/m2). We randomly assigned participants to either a behavioural intervention or standard antenatal care with an internet-based, computer-generated, randomisation procedure, minimising by age, ethnic origin, centre, BMI, and parity. The intervention was delivered once a week through eight health trainer-led sessions. Primary outcomes were gestational diabetes (diagnosed with an oral glucose tolerance test and by criteria from the International Association of Diabetes in Pregnancy Study Groups) and large-for-gestational-age infants (≥90th customised birthweight centile). Analysis was by intention to treat. This trial is registered with Current Controlled Trials, ISCRTN89971375. Recruitment and pregnancy outcomes are complete but childhood follow-up is ongoing.FindingsBetween March 31, 2009, and June 2, 2014, we assessed 8820 women for eligibility and recruited 1555, with a mean BMI of 36·3 kg/m2 (SD 4·8). 772 were randomly assigned to standard antenatal care and 783 were allocated the behavioural intervention, of which 651 and 629 women, respectively, completed an oral glucose tolerance test. Gestational diabetes was reported in 172 (26%) women in the standard care group compared with 160 (25%) in the intervention group (risk ratio 0·96, 95% CI 0·79–1·16; p=0·68). 61 (8%) of 751 babies in the standard care group were large for gestational age compared with 71 (9%) of 761 in the intervention group (1·15, 0·83–1·59; p=0·40). Thus, the primary outcomes did not differ between groups, despite improvements in some maternal secondary outcomes in the intervention group, including reduced dietary glycaemic load, gestational weight gain, and maternal sum-of-skinfold thicknesses, and increased physical activity. Adverse events included neonatal death (two in the standard care group and three in the intervention group) and fetal death in utero (ten in the standard care group and six in the intervention group). No maternal deaths were reported. Incidence of miscarriage (2% in the standard care group vs 2% in the intervention group), major obstetric haemorrhage (1% vs 3%), and small-for-gestational-age infants (≤5th customised birthweight centile; 6% vs 5%) did not differ between groups.InterpretationA behavioural intervention addressing diet and physical activity in women with obesity during pregnancy is not adequate to prevent gestational diabetes, or to reduce the incidence of large-for-gestational-age infants.FundingNational Institute for Health Research, Guys and St Thomas' Charity, Chief Scientist Office Scotland, Tommy's Charity.