This study evaluates the technical efficiency of Tunisia's health system over the period 1990–2022 and identifies the structural determinants that shape it. Existing assessments of Tunisian health performance rarely span the pre- and post-revolution periods within a single analytical framework, leaving policymakers without a longitudinal view of how systemic shocks reshaped efficiency. To address this gap, we apply a two-stage approach: Data Envelopment Analysis (DEA) under variable returns to scale (BCC model) to compute annual efficiency scores, followed by a Tobit regression to explain variation in those scores. The DEA specification uses three inputs (hospital beds per 1,000 people, number of inhabitants per doctor, and current health expenditure per capita in PPP) and two outputs (healthy life expectancy at birth and survival to age 65). Efficiency scores were computed with DEAP 2.1. The results reveal three distinct phases. During 1990–1999, efficiency scores remained above 0.7 in most years, reflecting a health system considered among the strongest in Africa. The 2000–2010 phase sustained high performance, with scores clustered near unity and reaching full efficiency in 2003. From 2011 onward, however, scores declined steadily from 0.829 to 0.601 in 2022, driven by post-revolution economic deterioration, brain drain among physicians, and weakening public investment. The Tobit estimation shows that the number of hospital beds is positively and significantly associated with efficiency, while poverty and population density exert significant negative effects. These findings suggest that expanding hospital capacity while addressing poverty and demographic pressure offers a viable path to restoring efficiency.