Background Increasing institutional births rates and improving access to comprehensive emergency obstetric care are central strategies for reducing maternal and neonatal deaths globally. While some studies show women consider service availability when determining where to deliver, the dynamics of how and why institutional birth rates change as comprehensive emergency obstetric care availability increases are unclear. Methods In this pre-post intervention study, we surveyed two exhaustive samples of postpartum women before and after comprehensive emergency obstetric care implementation at a hospital in rural Nepal. We developed a logistic regression model of institutional birth factors through manual backward selection of all significant covariates within and across periods. Qualitatively, we analyzed birth stories through immersion crystallization. Results Institutional birth rates increased after comprehensive emergency obstetric care implementation (from 30 to 77%, OR 7.7) at both hospital (OR 2.5) and low-level facilities (OR 4.6, p < 0.01 for all). The logistic regression indicated that comprehensive emergency obstetric care availability (OR 5.6), belief that the hospital is the safest birth location (OR 44.8), safety prioritization in decision-making (OR 7.7), and higher income (OR 1.1) predict institutional birth (p ≤ 0.01 for all). Qualitative analysis revealed comprehensive emergency obstetric care awareness, increased social expectation for institutional birth, and birth planning as important factors. Conclusion Comprehensive emergency obstetric care expansion appears to have generated significant demand for institutional births through increased safety perceptions and birth planning. Increasing comprehensive emergency obstetric care availability increases birth safety, but it may also be a mechanism for increasing the institutional birth rate in areas of under-utilization.