BACKGROUND: Herpes simplex keratitis (HSK) caused by herpes simplex virus 1 (HSV-1), which has high recurrent rate and incidence of severe vision loss, is the leading cause of infectious blindness in the world. The aim was to explore the clinical efficacy of oral ganciclovir (GCV) in the prevention of recurrent HSK. METHODS: A multicenter, prospective, randomized, single-blind, and controlled clinical trial was conducted from April 2010 to June 2013. One hundred seventy-three patients (173 eyes involved) who were diagnosed as recurrent HSK definitely, including stromal keratitis and corneal endotheliitis, were divided into three groups randomly: negative control (placebo) group was topically administered with 0.15% GCV ophthalmic gel, 4 times per day and 0.1% fluorometholone eye drops, 3 times per day until resolution of HSK; positive control acyclovir (ACV) group was topically adopted the same ophthalmic gel and eye drops and additionally received oral ACV 400 mg 5 times a day for 10 weeks and followed by 400 mg 2 times per day for 6 months; test GCV group was topically adopted the same treatment as negative control group and additionally received oral GCV 1000 mg 3 times per day for 8 weeks. The symptoms and signs were evaluated before and after the therapy 1 st week, 2 nd week and then followed up every 2 weeks until recovery. Furthermore, we followed up recurrence of HSK for every 3 months after recovery and then assessed the cure time, recurrent rate and adverse reactions. RESULTS: One hundred and seventy-three patients were followed up 7-48 months (mean 32.1 ± 12.3 months), but 34 patients were failed to follow-up. The cure time was 12.1 ± 4.3, 11.9 ± 4.0 weeks in negative control (placebo) group and positive control ACV group respectively (P = 0.991), which was longer than that in test GCV group (8.6 ± 2.8 weeks) and there was a significant difference between test GCV group and negative control (placebo) group or positive control ACV group (P = 0.000). Furthermore, the recurrent rate was higher in negative control (placebo) group (47.3%) than that in positive control group ACV (26.7%) and test GCV group (17.2%), and there was a great significant difference among the three groups (P = 0.007), but there was no significant difference between positive control ACV group and test GCV group (P = 0.358). In addition, there was no obvious adverse reaction expect neutropenia (only one patient in test GCV group). CONCLUSION: Short-term oral GCV could cure recurrent HSK and endotheliitis, shorten the course, reduce recurrent rate of HSK and have confirmed safety.