Background
Surgical site infection (SSI) following caesarean section is a serious but underreported problem with an estimated incidence of 5–9%. It is essential to identify adherence to established prevention strategies to reduce the incidence rate.

Aims
The aims of this study were to quantify unwarranted variation from evidence‐based practice on the prevention of SSI at caesarean section in Australia; and to identify predictors of not implementing an existing infection prevention bundle: pre‐incision antibiotic prophylaxis, vaginal preparation and spontaneous placenta removal.

Materials and methods
An online cross‐sectional survey of obstetricians and obstetric Diplomates was conducted in 2016. The primary outcome was adherence to an existing infection prevention bundle, with demographic and clinical variables predicting adherence through multivariable binary logistic regression.

Results
Forty‐nine percent of respondents (response rate 39.6%) reported implementing zero or only one element of the infection prevention bundle. The types of respondents most likely to have poor adherence were Diplomates (adjusted odds ratio (aOR) 2.58), obstetricians practising in private hospitals (aOR 3.34), those usually practising in public and private hospitals (aOR 2.23), and those not usually implementing a surgical safety checklist (aOR 3.77).

Conclusions
Adherence to best practice at caesarean section is low among many Australian obstetricians. Infection control practitioners and obstetricians need to collaboratively implement surgical safety checklists at caesarean section, and monitor implementation using process key performance indicators, and audit and feedback. These strategies will reduce unwarranted variation from evidence‐based infection control practice.