Objectives: to compare the patterns of a 17β-estradiol (E 2) gel containing 0.6 mg/g (1.5 mg E 2 per day, Gelestra); with the transdermal delivery system (Estraderm TTS 50) applied every 3 days over a 14-day period to women in spontaneous or surgical menopause. Methods: a single centre, open, randomised, parallel-group study was conducted. A total number of 33 postmenopausal women were enrolled. In 23 of them the menopause occurred spontaneously, while 10 women were bilaterally ovariectomized. Randomly, the subjects were treated with Estraderm TTS 50 (no. 8) or with Gelestra (no. 14). The pharmacokinetic study of the drugs was performed at the seventh, ninth and 14th day in Gelestra treated women and at the first, third and second day in Estraderm TTS 50 treated women. In fact, the seventh, ninth and 14th day of percutaneous treatment corresponds to the first, third and second day of application of the transdermal system application. Blood samples were taken by each subject at baseline and 1, 2, 3, 4, 8, 12 and 24 h after the gel or transdermal system application. In almost all samples the level of E 2 and estrone (E 1) were evaluated. Statistical analysis was performed by comparing the two groups of treatment. The following parameters were assessed: mean E 2 and E 1 concentrations, E 2 peak serum concentration within interval from 0 to 72 h ( C max), E 2 trough concentration within interval from 0 to 72 h ( C min), area under the E 2 time concentration curve in the interval from 0 to 72 h (AUC (0–72)), the average E 2 concentration during the measurement interval, calculated by dividing AUC (0–72) by 72 h ( C av), E 1/E 2 ratio, and percentage fluctuation (%Fluct) which is equal to 100 ( C max− C min/ C max). Results: there was no significant difference in E 2 C av between the two treatments. However, significant differences in favour to the gel on the first day (first h) and on third day (72nd h) and in favour to the patch at the second day (48th h) were detected. C max, E 1/E 2 ratio and AUC (0–72) were not statistically different, while a significantly higher C min for the gel was observed. Furthermore, the 90% confidence interval for AUC (0–72) ratio (0.83–1.10) was within the commonly applied bioequivalence acceptance range (0.80–1.25). The %Fluct was significantly lower for Gelestra than for Estraderm TTS 50. Conclusions: although the mean E 2 and E 1concentrations, C max, E 1/E 2 ratio and the AUC (0–72) did not differ between the two E 2 treatments, the Gelestra treatment showed a lower day-to-day variation over the three day application, than the Estraderm TTS 50.