Objective. To study the effect of hysterectomy or levonorgestrel-releasing intrauterine system (LNG-IUS) on premenstrual symptoms in women treated for menorrhagia. Design. Secondary analysis of a randomized controlled trial. Setting. Five university hospitals in Finland. Sample. A cohort of 236 women, aged 35–49 years (mean 43 years) referred for menorrhagia between 1994 and 1997. Women were not diagnosed with premenstrual syndrome. Methods. Women were randomized to treatment by hysterectomy (n=117) or LNG-IUS (n=119). Analyses were performed using the intention-to-treat and actual treatment principles. Women using estrogen therapy and women who underwent bilateral salpingo-oophorectomy were excluded from the analyses. Main outcome measures. The occurrence of premenstrual symptoms evaluated by questionnaires at baseline and at follow-up visits six and 12 months after the treatment and five years after the randomization. Results. Premenstrual symptoms decreased significantly in both groups by six months (p≤0.028) without significant differences between the groups, except that in the LNG-IUS group the decrease of breast tenderness was seen first by 12 months (p=0.048). Even though 42% of the women assigned to treatment with LNG-IUS were hysterectomized during the follow-up period, the results of intention-to-treat and actual treatment analyses were comparable. Conclusions. Both hysterectomy and LNG-IUS seem to alleviate premenstrual symptoms of women treated for menorrhagia, while the effect of these treatments on premenstrual syndrome remains unsettled.