This case–control study was designed to investigate the safety of the AID technology. The health status of the offspring conceived by 1620 couples who underwent 7272 AID cycles in our Center for Reproductive Medicine between June 2006 and December 2012 was retrospectively analysed. The control group included 1018 women who naturally conceived and delivered in the same period. Twin birth rate was significantly higher in the AID group (no triplet birth) than in the control group (2.01% versus 0.39%, P < 0.01). In the AID group, Caesarean delivery was used in 1299 cases (81.65%), spontaneous vaginal delivery (18.04%) and forceps-assisted vaginal delivery (0.31%).There was no significant difference in male/female ratio of the offspring between AID and control groups (113.55 : 100 versus 113.36 : 100, P > 0.05). Compared to natural pregnancy, a pregnancy through AID resulted in higher multiple birth rate, premature delivery rate and neonatal congenital malformation rate. Increased multiple birth rate was attributable to ovulation induction, and increased rate of low-birthweight infants was related to multiplets and premature delivery. Caesarean delivery was preferred in couples who received AID treatment. The male/female ratio of the AID offspring was similar between natural pregnancy and AID pregnancy.