To compare the efficacy and safety of two misoprostol dosing regimens for induction of labour in primiparous (1P) and multiparous (>1P) women. METHODS Retrospective study of induction of labour using vaginal misoprostol 25 μg vs. 50 μg every 6 h in 942 women at a tertiary centre. The main outcome variables are induction-to-delivery interval, latency period duration, vaginal delivery within 24 h, and maternal and foetal safety outcome. RESULTS With the 50 μg regimen, induction-to-delivery intervals were significantly shorter: 18.4 h vs. 24.6 h (1P) and 14 h vs. 17.9 h (>1P), as was latency period duration (by 5.4 and 4 h, respectively). Vaginal delivery within 24 h was significantly more frequent, as were non-reassuring foetal heart rate (1P: 20% vs. 14%) and tachysystole (1P: 31% vs. 11%; >1P: 21% vs. 7%). No uterine rupture was reported. Neonatal outcomes were similar except for significantly more frequent infant referral to neonatal intensive care in the >1P group receiving the 50 μg regimen (11% vs. 4%). CONCLUSION Vaginal misoprostol 25 μg seems to maintain efficacy with more acceptable maternal and neonatal safety. As induction of labour is an off-label use for misoprostol, safety should be prioritised with the lower dosage regimen despite the longer induction-to-delivery interval.