The dependencies of the internal and external diffusion limitations on the response and sensitivity of amperometric biosensors are investigated computationally using a two-compartment model, and a framework for evaluating and selecting the biosensor configuration is proposed. Enzyme-based biosensors are mathematically modelled by a system of reaction-diffusion equations containing a nonlinear term related to Michaelis-Menten kinetics. The biosensor response, sensitivity and stability are numerically analysed at the transition and steady state conditions in a wide range of model parameters, and attractive biosensor configurations are derived. The performed calculations show nonlinear effects of internal and external diffusion limitations on the biosensor current, response time and sensitivity as well as on the linear range of the calibration curve. The phenomenon of the non-monotonicity of the biosensor response is also investigated. (C) 2017 Elsevier Ltd. All rights reserved.