Novel classes of cannabinoid 2 receptor (CB2) agonists based on 1,2,3,4-tetrahydropyrrolo[3,4-b]indole and benzimidazole scaffolds have shown high binding affinity toward CB2 receptor and good selectivity over cannabinoid 1 receptor (CB1). A computational study of comparative molecular fields analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) was performed, initially on each series of agonists, and subsequently on all compounds together, in order to identify the key structural features impacting their binding affinity. The final CoMSIA model resulted to be the more predictive, showing cross-validated r2 (r(cv)2) = 0.680, non cross-validated r2 (r(ncv)2) = 0.97 and test set r²(r²(pred) = 0.93. The study provides useful suggestions for the design of new analogues with improved affinity.