Background Urinary catheterisation, which is associated with 80% of urinary tract infections (UTIs), is routinely performed prior to caesarean section without justification from the best available research evidence.Objectives To assess whether it is necessary to place indwelling urinary catheters routinely in caesarean section, and to determine the effects of this procedure on UTIs, urinary retention, intraoperative difficulties, operative complications, as well as other outcomes. Selection criteria Randomised controlled trials (RCTs) and nonrandomised controlled trials (NRCTs) comparing the use versus nonuse of indwelling urinary catheters in caesarean section were included.Data collection and analysis Two reviewers independently selected the studies and extracted the data. Results from the trials were combined to calculate relative risks (RRs) for dichotomous outcomes and mean differences (MDs) for continuous outcomes, with 95% confidence intervals (CIs).Main results Three trials (two RCTs and one NRCT) were included, involving a total of 1084 participants. Compared with the use of indwelling urinary catheters, nonuse had a significantly lower incidence of UTIs [RR 0.08; 95% CI 0.01, 0.64 (study design: RCT); RR 0.10; 95% CI 0.02, 0.57 (study design: NRCT)], a lower rate of discomfort at first voiding (RR 0.06; 95% CI 0.03, 0.12), less time until first voiding (MD -16.81; 95% CI -17.31, -16.31) and less time until ambulation (MD -6.01; 95% CI -6.68, -5.35); there were no statistically significant differences in the rate of urinary retention [RR 5.00; 95% CI 0.24, 103.18 (study design: RCT); RR 0.74; 95% CI 0.04, 15.18 (study design: NRCT)], operating time (MD -1.10; 95% CI -3.32, 1.12) and rate of intraoperative difficulties (RR 1.00; 95% CI -3.32, 1.12).Conclusions The nonuse of indwelling urinary catheters in caesarean section is associated with less UTIs and no increase in either urinary retention or intra-operative difficulties. Our results suggest that the routine use of indwelling urinary catheters for caesarean delivery in haemodynamically stable patients is not necessary, and can be harmful. However, better and larger randomised trials are needed to confirm these findings.