Background An index of the need for health visiting in general practice populations in the United Kingdom was developed, using routinely held data, to inform decisions about the deployment of health visitors. Methods The following indicators of need for health visiting were developed by consensus among health visitors and others: the population aged under 5 years; elective admissions under 5 years; births under 2500 g; deaths under 65 years; all expressed as rates per 10000 people registered with general practices irv Norfolk. All indicators were compared with the number of health visitors per 10000 people, obtained by a postal survey of health visitors. The indicators were converted to Z-scores and summed to produce a composite score of need for each general practice. The results were compared with the results of a workload profile using data compiled by health visitors within one Primary Care Group. Results Health visitors are not allocated according to need at either the practice or Primary Care Group level. The Pearson's correlation coefficient between the allocation suggested by this method and current allocation is 0.37 (p < 0.01). The correlation between this method and the allocation suggested by health visitors' workload profiling in one Primary Care Group was 0.76 (p < 0.01 ). Conclusions Health visitors are currently distributed according to historic patterns rather than need. This paper describes a simple method of determining need at general practice level, which can be used to allocate health visitors equitably.