INTRODUCTION: Social media platforms are used by over 4.9 billion people for networking and community building, as well as for healthcare information seeking and decision-making. Most studies investigating the utilization of social media during pregnancy have focused on high-income countries, despite the surge in social media utilization globally. In this study, we analyzed how pregnant women in Ghana, West Africa, utilize Facebook to inform decision-making on their most salient pregnancy concerns. METHODS: We utilized machine learning techniques (Web Scraping and Latent Dirichlet Allocation) to mine and analyze posts from the Ghana-based MidWife Sally Pregnancy School Facebook group between August 16, 2020 and April 29, 2023. Posts were extracted, cleaned, and analyzed using Gensim python library. Topics were generated based on their probabilities and relevance to the study goal. RESULTS: A total of 3,328 posts were extracted and 3,322 were analyzed after removing 6 empty posts. Seven major topics with listed subtopics were identified: Pregnant (693 posts): personal physiological changes, exercise during pregnancy, medication (e.g., anti-malarials, pain killers) Delivery (367): emergency delivery, vaginal/caesarean birthing, breastmilk production, exercise during pregnancy Pain (350): location of pain and pain relief modalities (e.g., exercise, medication, sleep) Breastfeeding (248): delivery, emergency service, milk production Water (174): cold water consumption, infant feeding (e.g., gripe water, constipation, formula) Sleeping (165): discomfort, sleeping positions, exercise to induce sleep, sleep as a natural analgesic Antenatal (124): fetal growth, progress, hospital selection Of note, content from "Pregnant", "Delivery" and "Sleeping" included mentions of depression, while "Breastfeeding" highlighted cultural approaches to increasing milk production. The sentiment analysis showed that 43.4% of the responses were neutral and primarily focused on seeking information. Negative sentiments, which were more distressing, comprised 46.4% of the responses, while positive sentiments, had a celebratory tone and represented 10.2% of the data. CONCLUSION: Social media analysis, previously employed in high income settings, can provide impactful, granular snapshots of pregnant people's concerns in the African region, which could be used to inform social media interventions aimed at filling educational gaps in antenatal care for those without adequate healthcare access.