Background The Department of Health report on neonatal care in England (2003) supported the British Association of Perinatal Medicine standards defining requirements of babies with different clinical dependency and recommended that neonatal units (NNU) be organised into networks. Each of the 24 networks has designated its constituent units locally but in a non-standardised way.

Objective To develop a standard method to categorise NNU on the basis of activity and staffing that can be used to examine whether clinical outcomes for extremely preterm babies are related to NNU characteristics.

Methods Clinical activity and consultant staffing in 2005 were collected for English NNU. Units were categorised to the highest level satisfied by both criteria agreed by consensus by the study team. The results were compared with local designations of levels 3, 2 and 1 supplied by network clinical leads.

Results There was an overall agreement of level of care for 110 (60.4%) out of the total 182 English NNU. Only 25 of the 46 NNU designated as level 3 by networks achieved the study criteria for high activity units; eight networks did not include a “high activity” unit (see table).

Conclusion There is no agreement as to the optimal size and senior staffing for NNU; the scheme proposed by the study team does not map well onto those designations currently used by the networks.