A series of 20 compounds isolated from some pyrazole derivatives were subjected to cytotoxicity test against HIV–1. Two statistical approaches namely: Genetic Function Algorithm (GFA) and Multi Linear Regression Analysis (MLRA) were effectively used. Best three descriptors which are: VR2_Dzv, VR1_Dzp and PubchemFP824 were selected for the Quantitative structural and activity relationship (QSAR) using the two aforementioned statistical approaches. The results obtained were as follows: R-squared (R2) of 0.9698, adjusted squared (R2adj) of 0.9607, cross validated R-squared (LOO- Q2cv) value of 0.9299 and external prediction ability (R2pred) of 0.6827. The result proved that the compounds are attractive platform and clinically viable for developing anti HIV-1 drugs. Multivariate statistics with chemical descriptors molecular shape and polarizability may be useful for the evaluation of cytotoxicity of pyrazole.