In this study, multilayer perceptrons (MLPs) and conventional isotherm equations were used to model bisorption of nickel using terrestrial moss. Characterization of the biosorbent was examined by scanning electron microscopy and Fourier transform infrared spectroscopy analyses. Batch biosorption tests were performed to examine the impacts of different experimental conditions. Thermodynamic calculations were made to evaluate the feasibility of the biosorption. All of the experimental data (total 86 data-sets) were used for MLP modeling purposes whereas equilibrium data were used in Langmuir, Freundlich, and Temkin models. Adsorption kinetic data were tested using pseudo-first-order and second-order kinetic models. Performances of the models were evaluated considering calculated R-2 and mean standard error (MSE) values. Related isotherm and kinetic parameters were also calculated for conventional biosorption equations. In multilayer perceptron (MLP) modeling studies, network architecture with three hidden layers provided highest prediction efficiency. Although both MLPs and conventional models are regarded to be useful, perceptron models are thought to provide more representative information as all factors are evaluated simultaneously.