Background: Despite continued interest in the use of mobile health for improving maternal health outcomes, there have been limited attempts to identify relevant program theories. Objectives: This study had two aims: first, to explicate the assumptions of program designers, which we call the program theory and second, to contrast this program theory with empirical data to gain a better understanding of mechanisms, facilitators, and barriers related to the program outcomes. Methods: To achieve the aforementioned objectives, we conducted a retrospective qualitative study of a text messaging (short message service) platform geared at improving individual maternal health outcomes in Uganda. Through interviews with program designers (n=3), we elicited 3 main designers’ assumptions and explored these against data from qualitative interviews with primary beneficiaries (n=26; 15 women and 11 men) and health service providers (n=6), as well as 6 focus group discussions with village health team members (n=50) who were all involved in the program. Results: Our study results highlighted that while the program designers’ assumptions were appropriate, additional mechanisms and contextual factors, such as the importance of incentives for village health team members, mobile phone ownership, and health system factors should have been considered. Conclusions: Our results indicate that text messages could be an effective part of a more comprehensive maternal health program when context and system barriers are identified and addressed in the program theories.