Optimal nutrition is one of the fundamental components for infants to reach their full growth and neurodevelopmental potential. Best practice is facilitated by a contemporaneous, multidisciplinary, evidence-based nutrition policy. Such evidence has recently been reviewed. We have assessed: the prevalence of nutrition policies in neonatal units in the UK and Eire; their application to hypothetical cases; the availability of dietetic input; and whether any differences existed between non-regional and regional units. A standardized questionnaire was devised by a multidisciplinary group and posted to all 255 neonatal units in the UK and Eire in 2002. Replies from 67 neonatal units were received: 48 out of 233 non-regional and 19 out of 22 regional units. A feeding policy was present in 33 units, and regular access to dietitians occurred in 37 units. For a hypothetical infant less than 28 weeks' gestation, enteral feeds would be commenced at 0-2 days in 81% of non-regional and 94% of regional units (P = ns), and be continuous in 11% of non-regional and 32% of regional units, and bolus feeding in 89% of non-regional and 68% of regional units (P = ns). Routine fortification of breastmilk would occur more frequently in non-regional units (96%) than in regional units (79%) (P = 0.050). Vitamin and iron supplements would be given to infants receiving postdischarge or high-energy milks in 68% of non-regional units and in 79% of regional units (P = ns). Calorie counts (63% regional vs. 8% non-regional, P < 0.001), and daily weights (68% regional vs. 33% non-regional, P = 0.014), were used more frequently in regional units. Many units surveyed did not have a nutrition policy. Many infants receive unnecessary additional vitamins and supplements. Practice is variable throughout the country, but we found no evidence of major differences between regional and non-regional units, apart from their monitoring of growth and rates of breastmilk fortifier usage.