Jumonji domain-containing protein 2 (JMJD2C) plays a role in increasing the production of Mdm2 oncogene in overproduction condition. Further, it will inhibit the cell cycle of p53 to suppress tumor growth. The purpose of this research was to study Quantitative Structure Activity Relationship (QSAR), design new compounds, predict their toxicity, and interaction of new compounds with the receptor. Optimization of three-dimensional structure was performed using HyperChem 8.0.3 with Semi-empirical AM1 method with iteration limit of 50 and convergence limit of 0.01. Leave One Out Cross Validation (LOO-CV) was done for predictor value calculation and determination of outlier compounds using MOE 2007.09. Multilinear regression analysis was conducted by SPSS Statistics 17.0. The best QSAR equation was used to find new compounds with better activity than the parent compound. Study of toxicity of new compounds to humans was performed with Toxtree 2.1.0 and that to aquatic organisms was obtained with Ecosar 1.00a. The interaction of new compounds with receptors was observed using Argus Lab 4.0.1 and AutoDock Vina. The best QSAR equation was: IC50 = -3914.398(±369.478) - 277.686(±29.759) AM1_HOMO - 176.811(±18.602) logS + 2.820(±0.206)ASA_H - 368.008(±20.701) logP(o/w). Three new compounds were obtained with better activity than that of the parent compound. Compound 1 and 3 were potential candidates for better interaction with receptors JMJD2C.