OBJECTIVES: To investigate the cost effectiveness of intermittent treatment of PMDD (premenstrual dysphoric disorder) with Premalex (escitalopram) 20 mg compared with sertraline 50-100 mg based upon the Clinical Global Impressions – Severity (CGI-S), from a societal perspective. METHODS: We identified one randomised placebo controlled trial with sertraline and one with Premalex, reporting the CGI-S as an outcome. Using placebo, the CGI-S was used to make an indirect effect comparison between Premalex and sertraline. The CGI-S was translated into QALY weights, through the proportion of time spent with a high degree of the anxiety/depression in the EQ-5D. Costs of health care visits were estimated using a local treatment pattern survey among GPs and gynaecologists. Official drug prices were used. A Premalex dose of 15 mg, the average of 10 and 20 mg, was assumed as it is stated in the SPC text that many of the patients will benefit from 10 mg. 37-75% lower drug costs were used in years 2-3 due to expected generic competition. Indirect costs were estimated using a published international study of the effect of PMDD on sick leave and productivity. A societal perspective was taken over a 3-year time frame. RESULTS: During the first year, Premalex treatment increased drug costs (SEK 1599), partly offset by indirect costs saving (SEK 1413), resulting in a total cost of SEK 186, compared to sertraline. An estimated gain of 0.0044 QALYs with Premalex compared to sertraline gave an incremental cost per QALY gained of SEK 42200 (EUR 4700). In the next 2 years, drug costs were more than offset by indirect costs saving, leading to an overall gain of 0.0132 QALYs and a savings of SEK 1600 per three years with Premalex compared to sertraline. CONCLUSIONS: Treatment of PMDD with Premalex is cost effective compared to sertraline, from a societal perspective.