Via a five-step-reaction procedure for the preparation of 19 known N-[2-chloro-9-(tetrahydropyran-2-yl)-9H-purin-6-yl]-N-cyclopropylglycylamino acid benzylesters (6a-s) and successive removal of 9-(tetrahydropyran-2-yl) and benzylester groups 19 novel N-[2-chloro-9H-purin-6-yl]-N-cyclopropylglycylamino acid benzylesters (7a-s) and 19 novel N-[2-chloro-9H-purin-6-yl]-N-cyclopropylglycylamino acids (8a-s) were provided. On tail-flick mouse model the in vivo analgesic activities of these 38 novel compounds were measured and most of them were defined as good analgesics. Based on Molecular Field Analysis of the pain threshold variations of the mice receiving 48 compounds in terms of the descriptors proton and methyl an equation was established. The data points (n), correlation coefficient (r), and square correlation coefficient (r(2)) of this equation were 48, 0.923, and 0.852, respectively. Using this equation pain threshold variations of 9 compounds were predicted and the errors ranged from 1.71 to 8.92.