P21-activated kinase 4 (PAK4) is a serine/threonine protein kinase, which is associated with many cancer diseases, and thus being considered as a potential drug target. In this study, three-dimensional quantitative structure-activity relationship (3D-QSAR), molecular docking and molecular dynamics (MD) simulations were performed to explore the structure-activity relationship of a series of pyrropyrazole PAK4 inhibitors. The statistical parameters of comparative molecular field analysis (CoMFA, Q 2 = 0.837, R 2 = 0.990, and R 2 pred = 0.967) and comparative molecular similarity indices analysis (CoMSIA, Q 2 = 0.720, R 2 = 0.972, and R 2 pred = 0.946) were obtained from 3D-QSAR model, which exhibited good predictive ability and significant statistical reliability. The binding mode of PAK4 with its inhibitors was obtained through molecular docking study, which indicated that the residues of GLU396, LEU398, LYS350, and ASP458 were important for activity. Molecular mechanics generalized born surface area (MM-GBSA) method was performed to calculate the binding free energy, which indicated that the coulomb, lipophilic and van der Waals (vdW) interactions made major contributions to the binding affinity. Furthermore, through 100 ns MD simulations, we obtained the key amino acid residues and the types of interactions they participated in. Based on the constructed 3D-QSAR model, some novel pyrropyrazole derivatives targeting PAK4 were designed with improved predicted activities. Pharmacokinetic and toxicity predictions of the designed PAK4 inhibitors were obtained by the pkCSM, indicating these compounds had better absorption, distribution, metabolism, excretion and toxicity (ADMET) properties. Above research provided a valuable insight for developing novel and effective pyrropyrazole compounds targeting PAK4.