In response to the disappointing outcome data of the population-based EPICure study published in 2000, we compared the outcome of infants 22 0/7 to 25 6/7 weeks of gestational age (GA) in a single tertiary care centre 2000-2004 with that of EPICure. EPICure tools and definitions, including 30 months' Bayley Scales. Of 83 infants <26 weeks born alive, more were admitted to intensive care--82% vs. 68% (p < 0.0001)--and more infants survived to discharge (57% vs. 26%, p < 0.0001; 69% vs. 39%, p < 0.01, of those admitted to intensive care). More infants, as a percentage of live births, survived without severe (41%, 34/83 vs. 20%, 233/1185, p < 0.0001) or overall disability (22%, 18/83 vs. 13%, 155/1185, p = 0.03). However, at the border of viability--GA 23 and 24 weeks--the rate of infants surviving without overall disability was not significantly higher (13%, 6/45 vs. 9%, 56/623). In infants <26 weeks of GA, increased rates of survival and survival without disability were observed in a single-centre inborn cohort born 5-8 years later than the EPICure cohort. This did not translate into increased survival without overall disability in infants of 23-24 weeks of GA.