This prospective study was undertaken to investigate the efficacy of nasal continuous positive airway pressure (nCPAP) in a group of extremely small infants denied access to a neonatal intensive care unit (NICU) in South Africa. Consecutive infants weighing less than 1200 g and/or of a gestational age below 28 weeks admitted to the neonatal ward with respiratory distress syndrome (RDS), and who were refused admission to the NICU, received either nCPAP (Infant Flow System E.M.E., UK) of headbox oxygen. Of 22 infants, 11 infants were included in the treatment group (nCPAP) and 10 in the control group. Within the first 24 h, two infants (18 per cent) in the nCPAP group and eight infants (80 per cent) in the control group died (p = 0.007) (survival OR = 18; RR = 4.09). A statistically significant improvement in the arterial-alveolar (a/A) oxygen ratio occurred in the nCPAP group between postnatal day 1 and day 3 of life (0.17 vs. 0.36; p < 0.005). Neonatal complications occurred in six (55 per cent) infants who survived the first 24 h of life. Eighty per cent of the infants with intraventricular haemorrhage (IVH) died, as well as all the infants who were born before arrival at the hospital. At the time of discharge from hospital, 45 per cent (five infants) in the nCPAP group survived vs. 20 per cent (two infants) in the control group. The neurodevelopmental outcome of six of the surviving seven infants were evaluated at 1 year of corrected age. The neurodevelopmental outcome as assessed by the Griffith Score was within normal limits in all infants. One infant has sensorineural deafness and one is deaf and has a possible mild spastic diplegia (both in the treatment group). We conclude that nCPAP significantly improves the short-term survival of very low birth-weight (VLBW) infants with moderate to severe respiratory distress syndrome who could not be admitted to intensive care. nCPAP significantly improves the a/A oxygen ratio between day 1 and day 3 of life.