Poor health worker performance is a well-documented obstacle to quality service provision. Due to the increasingly widespread availability of mobile devices, mobile health (mHealth) has received growing attention as a service improvement tool. This pilot study explored feasibility, acceptability and outcomes of an mHealth intervention designed to increase coverage of intermittent preventive treatment of malaria in pregnancy (IPTp) in two districts of West Nile, Uganda. In both districts, selected health workers (N = 48) received classroom training on malaria in pregnancy. All health workers in one district (N = 49) subsequently received 24 text messages reinforcing the training content. The intervention was evaluated using a mixed-methods approach, including four focus group discussions with health workers and three in-depth interviews with district health officials, health worker knowledge assessments one month (N = 90) and six months (N = 89) after the classroom training, and calculation of IPTp coverage from participating health facilities' (N = 16) antenatal care registers covering six months pre- and post-intervention. Complementing classroom training with text messaging was found to be a feasible, acceptable and inexpensive approach to improving health worker performance. The messages served as reminders to those who had attended the classroom training and helped spread information to those who had not. Health workers in the district where text messages were sent had significantly better knowledge of IPTp, achieving an increased composite knowledge score of 6.00 points (maximum score: 40) compared with those in the district where only classroom training was provided. Average facility coverage of three doses of IPTp was also significantly higher where text messages were sent (85.8%) compared with the district where only classroom training was provided (54.1%). This intervention shows promise for the improvement of health worker performance for delivery of IPTp, and could have significant broader application.