OBJECTIVE: To investigate the effect of a two-component intervention to change hospital practice with regard to the timing of umbilical cord clamping. METHODS: A pre-/post-study design was used to measure the effect of a two-component intervention on mean time to clamp the umbilical cord. The study took place at Hospital Iquitos "César Garayar García" in Iquitos, Peru. A total of 224 women were recruited from the hospital labor room: 112 pre-intervention, from 18 May-3 June 2009, and 112 post-intervention, from 6-20 July 2009. The intervention consisted of 1) a "best practice" three-day training workshop on birthing, and 2) a hospital directive. All deliveries were observed and the time between delivery of the first shoulder and clamping of the umbilical cord was measured with a digital stopwatch. RESULTS: The mean time between delivery and cord clamping before the intervention was 56.8 seconds (95% confidence interval [CI]: 51.0, 62.7). This increased to 169.8 seconds (95% CI: 153.8, 185.8) following the intervention. The difference in mean time to clamp remained significant in multivariate analyses (βadjusted = 113.2 seconds, 95% CI: 96.6, 129.9). CONCLUSIONS: Hospital policy and practice can be successfully changed from early to delayed umbilical cord clamping using a simple, two-component intervention.