OBJECTIVES: Patients access to new drugs is particularly a challenging task in UK. English Primary Care Trusts (PCT) play a key role in facilitating the new drug’s entry. The PCTs differ from each others through their behaviors and characteristics. The main purpose of this study is to bifurcate a total of 152 PCT in similar clusters having homogenous behavior, applying several clustering approaches, and to compare the different approaches METHODS: All available information about the PCTS is obtained from public sources and one key+ data base. A total of 64 variables were identified and classified into 5 groups according to population’s profile, prescription and patient’s characteristics, economic, financial as well as organizational criteria. We applied three kinds of clustering approaches: the two first, namely k-means (implemented both directly and indirectly) and Hierarchical Agglomerative Clustering (HAC) are traditional ones while the third, namely Self Organizing Map (SOM) is a variant of neural network architectures. Finally, contingency analysis and khi-square statistics were considered to encircle both dependencies and similarities between clusters RESULTS: Preliminary estimates have identified five clusters (for each method) that are different in their intrinsic characteristics. Contingency results reflect strong dependence between the direct k-mean and both HAC and SOM: indeed clusters are not very widely scattered between them. The chi-square (p-value) test corroborates their homogeneity. The indirect K-means reflects heterogeneity in cluster’s dispersions also confirmed by the chi-square test CONCLUSIONS: Despite the fact that clusters are generated from different approaches and have no similar characteristics, a plausible correspondence seems to develop between them. Thus, our investigations may help in identifying countless promising market strategies facilitating the access of new drugs according to UK regional standards