BACKGROUND: Levonorgestrel (LNG) consistently prevents follicular rupture only when it is given before the onset of the ovulatory stimulus. As locally synthesized prostaglandin (PG) plays a crucial role in follicular rupture and cyclooxygenase-2 (cox-2) catalyses the final step of PG synthesis, we reasoned that adding a cox-2 inhibitor to LNG would prevent follicular rupture even after the ovulatory process had been triggered by the gonadotrophin surge. METHODS: Forty-one women were divided into two groups. One was treated when the size of the leading follicle was 15–17 mm (n = 10) and the other when it was ≥18 mm (n = 31). Each woman contributed with one cycle treated with LNG 1.5 mg single dose plus placebo and another treated with LNG + meloxicam (Melox) 15 mg, in a randomized order. Serial blood sampling for the assay of LH and follicular monitoring by transvaginal ultrasound were performed before and after treatment. RESULTS: Follicular rupture failed to occur within the 5-day period that followed treatment in 50 and 70% of cycles treated with LNG + Placebo and LNG + Melox, respectively, in the 15–17 mm group (P = 0.15) and in 16 and 39% of cycles treated with LNG + Placebo and LNG + Melox, respectively, in the ≥18 mm group (P < 0.052). The overall proportion of cycles with no follicular rupture or ovulatory dysfunction increased significantly by the addition of Melox to LNG (66 versus 88%, P < 0.012; n = 41-matched pairs). CONCLUSIONS: The trend towards increased incidence of no follicular rupture when Melox was combined with LNG suggests that the addition of a cox-2 inhibitor has the potential to improve the contraceptive efficacy of LNG by a pre-fertilization effect.