OBJECTIVES: To assess the cost-effectiveness of travoprost versus latanoprost as single agents to treat glaucoma in France. METHODS: A Markov model reproduced the course, over 5 years, of patients beginning a prostaglandin as monotherapy (PM). The effectiveness criterion was ‘mean time to treatment change' (MTTC), fitted with a Weibull distribution from a national survey. Possible switches were association (A), treatment substitution (TS) and laser treatment or surgery (LS). After LS, patients could remain without treatment or proceed to PM or TS. Stratification used intra-ocular pressure (IOP) at treatment onset: ≤20 mmHg, 21 to 23 mmHg and ≥24 mmHg. Transition probabilities and costs per treatment line were extracted from two French observational databases. Bootstrap techniques were implemented to drive the probabilistic sensitivity analyses. RESULTS: MTTC was 44.3 months for travoprost and 37.7 for latanoprost. Additional costs for Travatan were €52, leading to an ‘incremental cost-effectiveness ratio' (ICER) without treatment change of €95 per year. 1.9% of patients treated with latanoprost underwent laser treatment or surgery, compared to 1.2% with travoprost. Results varied with baseline IOP values (≤20, 21 to 23, ≥24 mmHg) such that 55.6%, 53.9% and 50.4% of patients, respectively, remained under travoprost treatment when simulation ended, compared to 32.3%, 26.1% and 26.1% under latanoprost. Thus ICERs, without treatment change, were €140, €45, and €123 per year, respectively. CONCLUSION: Travoprost yielded a longer effectiveness profile and minimized early treatment regimen changes. The smaller portion of patients needing a new treatment, laser treatment or surgery virtually compensated for the higher travoprost acquisition cost. Travoprost is a more cost-effective alternative, especially in patients whose IOP at treatment onset lay between 21 and 23 mmHg.