OBJECTIVE: To determine the associations between maternal ethnicity and outcomes of infants born between 22 and 31 weeks' gestation and admitted to neonatal intensive care units in New South Wales and the Australian Capital Territory, Australia, between 1995 and 2006. DESIGN AND PATIENTS: De-identified perinatal and neonatal outcome data for 10 267 infants were examined. There were 8629 (84.0%) Caucasian, 922 (9.0%) Asian, 439 (4.3%) indigenous, 127 (1.2%) Polynesian and Maori (PAM) and 150 (1.5%) infants of other maternal ethnicities (excluded from study). Caucasians were the referent for all comparisons. RESULTS: Infants of indigenous mothers were less likely to receive antenatal steroids and three times as likely to be born in non-tertiary hospitals (OR 3.28, 95% CI 2.59 to 4.16, p<0.001). PAM infants were more likely to have Apgar scores <7 at 5 min of age (1.76, 95% CI 1.16 to 2.67, p<0.01). Asian infants had lower birth weight (mean±SD 44.7±27.9, p<0.001) and head circumference percentiles (47.8±29.0, p<0.001), were more likely to be small for gestational age (1.53, 95% CI 1.25 to 1.88, p<0.001), less likely to have hyaline membrane disease (0.78, 95% CI 0.68 to 0.90, p<0.001) but had a higher risk of severe retinopathy of prematurity (1.52, 95% CI 1.11 to 2.07, p<0.01). Ethnicity did not influence infant mortality. CONCLUSIONS: Neonatal growth characteristics and morbidity but not mortality are influenced by maternal ethnicity. Of concern is the risk of low Apgar scores in PAM infants and non-tertiary births of indigenous infants. Review of perinatal care for certain vulnerable ethnic populations is recommended due to the rapidly changing ethnic compositions of many countries around the world.