AIM: Sudden unexpected death in infancy (SUDI) rates for Maori and Pacific infants remain higher than for other ethnic groups in New Zealand and bed-sharing is a major risk factor when there is smoking exposure in pregnancy. Sleep space programmes of education and Pepi-Pod baby beds require evaluation. METHODS: Two hundred and forty Maori and Pacific women and infants were randomised 1:1, to the Pepi-Pod sleep space programme, or to a control group with ‘usual care’. When infants were under 2 weeks of age, baseline interviews occurred, followed up by interviews at 2 and 4 months of age to assess safe sleep knowledge, infant care practices and Pepi-Pod use and acceptability. All participants were offered a New Zealand Standard approved portable cot. RESULTS: At baseline, 25% of babies did not have a baby bed. Knowledge of smoking and bed-sharing as SUDI risks improved at follow-up in both groups. One quarter regularly bed-shared at follow-up in both groups. Intention to bed-share was a strong predictor of subsequent behaviour. Pepi-Pods were regularly used by 46% at 2 months and 16% at 4 months follow-up. CONCLUSIONS: Bed-sharing and knowledge improvement were similar irrespective of group. It is likely that the impact of the intervention was reduced because the control group received better support than ‘usual care’ and all participants had a baby bed. New Zealand SUDI rates have declined since sleep space programmes have been available. Sleep space programmes should be prioritised for those with modifiable SUDI risk.