Cyclosporine has shown promising results for mortality in patients with Stevens-Johnson syndrome/toxic epidermal necrolysis. However, available studies included only a small number of patients and did not include a validated and homogenous control group. We present the results from a retrospective monocentric study including 174 patients with Stevens-Johnson syndrome/toxic epidermal necrolysis during 2005–2016. Among them, 95 received cyclosporine (3 mg/kg/day) plus supportive care, and 79 received supportive care only. Both a traditional exposed/unexposed method and a propensity score-matching method were used to compare the progression of skin detachment between day 0 and day 5, the proportion of patients with cutaneous re-epithelialization starting on day 5 or mucosal re-epithelialization on day 10, the duration of progression, and the number of deaths between the two groups. None of these outcomes significantly favored cyclosporine, either by the exposed/unexposed method or the propensity score method. Acute renal failure affected more patients receiving cyclosporine (P = 0.05). Overall, the results of this epidemiological study did not show a beneficial effect of cyclosporine in patients with Stevens-Johnson syndrome/toxic epidermal necrolysis. They are discordant with those previously published. The large number of patients and the use of a propensity score method provide valuable insights. The main limitation of the study is the lack of randomization.