Adsorption performance, multivariate interaction mechanism and sensitivity analysis for the removal of the herbicide metolachlor (MET) by MIL-101(Cr) metal organic framework (MOF) are investigated using experimental, optimization models and computational technique. The MOF adsorbent was hydrothermally synthesized and characterized by Powdered X-ray diffraction (XRD), Fourier Transformed Infrared (FTIR) and thermogravimetric analysis (TGA). The adsorption kinetics and isotherm models show the presence of multilayer adsorption and chemisorption controlled adsorption process. High adsorption capacity of 238.041 mg/g was recorded with fast equilibration time of ~ 30 min, R2 0.999 and AIC �50.655. The thermodynamic parameters follow an endothermic and spontaneous adsorption with �H° 32.872 kJ/mol. The response surface methodology (RSM) was used to design the minimum significant number of experimental runs (44 runs) showing simultaneous multivariate interaction of process parameters with R2 0 990. The model F-value of 122.45 show that the model is well correlated. The artificial neural network learning algorithm was employed to predict the adsorption of MET with high level of accuracy R2 0.999 and RMSE 0.047. The best prediction architecture was obtained using the 5�10-1 topology. The predicted values of the RSM and ANN are highly correlated with the experimental results. The docking simulation was used to study the interaction between the MOF and the pollutant. Prospects for the MOF to be used repeatedly was also evaluated for six cycles and the MOF still maintain over 90 removal efficiency. The findings show the potential of the MOF for the effective remediation of MET in aqueous medium. © 2022, The Author(s) under exclusive licence to Iranian Society of Environmentalists (IRSEN) and Science and Research Branch, Islamic Azad University.