Background: Today, scientific evidence has supported the popular belief that physical activity is associated with biological health in pregnant women. A randomized controlled trial was used to assess the benefits of physical exercise during pregnancy on maternal lipids in low-income Latina women. Methods: The study included 67 nulliparous low-income Latina women in gestational weeks 16-20, randomly assigned into one of two groups: 1) The exercise group, which took part in aerobic and resistance exercise for 60min, three times a week for 12weeks, 2) The control group, which undertook their usual physical activity and prenatal care. The primary outcomes were changes in maternal blood lipids after intervention. Obstetrical and neonatal outcomes measured were type of delivery, postpartum hemorrhage, newborn and/or maternal complications', gestational age, weight gain, birth weight, foetal growth, and Apgar score. Results: Fifty women completed the study. At the end of the intervention, there were differences between groups in low-density lipoprotein levels (mean change: -8mg/dL, 95%CI -3 to -29; P < 0.001) and triglycerides (mean change: -6mg/dL, 95%CI -1 to -11; P=0.03). Also, compared with women who remained in the control group, active women showed lower complications during delivery (moderate postpartum haemorrhage) (58% compared with 75%; P < 0.05) and lower complications in newborns (e.g. cyanosis or respiratory distress) (21% compared with 46%; P < 0.001). Conclusions: An exercise programme during the second and third trimester favours less gain in low-density lipoprotein cholesterol and triglycerides fewer delivery and neonatal complications. Trial registration:NCT00741312(August 22, 2008).