OBJECTIVES: The process of disinvestment from inefficient healthcare involves identification and prioritisation of candidates, a health technology reassessment (HTR)of evidence, implementation and monitoring of discontinuance.  We evaluate whether variation in procedure rates is a useful tool for identifying potential candidates for HTR and disinvestment. METHODS: We used English Hospital Episode Statistics (HES) data to identify inpatient procedures.  We selected the 181 most frequent interventional procedures for analysis. For each procedure we used Poisson regression to estimate the variance in procedure rates, adjusting for age, gender and other proxies of clinical need, between Primary Care Trusts in England.  We conducted multivariate regression analyses to examine factors that might be associated with high variation in procedure rates (e.g. coding uncertainty, evolving evidence).   RESULTS: The degree of inter-PCT variation in procedure rates differed vastly from procedure to procedure. Among the five procedures with the highest inter-PCT variance, the procedure rate was more than thirty times higher in the PCT at the ninetieth percentile than the PCT at the tenth percentile.  The multivariable analysis provided strong evidence that large increases in procedure use, large decreases in procedure use, the presence of a substitute procedure, and shorter length of stay were all associated with higher inter-PCT variation in procedure rates.  CONCLUSIONS: The widespread geographic variation in hospital procedure rates in England are not solely due to variance in clinical need and are likely to reflect clinical uncertainty about appropriate procedure use which might be reduced by HTR.  The relevant HTR questions often concern the appropriate procedure setting and patient subgroups or the relative value of two alternative procedures rather than the value of a single procedure per se.  In some circumstances knowledge of geographic variation might lead to NHS savings and disinvestment or discontinuation of inefficiently used procedures.