background: This is a prospective long-term extension study of a randomized controlled trial aimed to assess the risk–benefit ratio of an ultra-conservative fertility-sparing approach in patients with bilateral borderline ovarian tumours (BOTs). methods: The experimental group (n ¼ 15) was treated with an ultra-conservative surgical approach consisting of bilateral cystectomy, whereas the control group (n ¼ 17) received a less conservative surgery consisting of oophorectomy plus controlateral cystectomy alone. All patients received a complete laparoscopic staging followed by a fertility enhancement programme. Patients who completed childbearing were treated with a non-conservative standard treatment at the first recurrence. results: After a follow-up period of 128 (9 interquartile range (IQR); 115–150 range) and 132 (7 IQR; 117– 152 range) months for the experimental and control groups, respectively (P ¼ 0.25), the time to first baby-in-arm (P , 0.02) and the relative rate (RR) of baby-in-arm (8.05 [95% confidence interval (CI), 1.20 –9.66; P , 0.01]) were significantly lower and higher, respectively, for the experimental compared with the control group. Although the time to first recurrence was significently (P , 0.01) shorter for the experimental group, in the regression analysis the difference did not reach the statistic significance (P ¼ 0.14), and the RR of recurrence (1.23 [95% CI, 0.62 –3.17; P ¼ 0.41]) was not significant. Finally the number needed to treat for pregnancy was three, the number needed to harm for radical surgery was only two. conclusions: The ultra-conservative fertility-sparing approach is more effective than the standard approach in terms of reproductive outcomes, but presents a higher oncological risk.