In this work, we investigate the effects of various types of valve behavior (e.g., linear valve dynamics and stiction) on the effectiveness of process control in a unified framework based on systems of nonlinear ordinary differential equations that characterize the dynamics of closed-loop systems including the process, valve, and controller dynamics. By analyzing the resulting dynamic models, we demonstrate that the responses of the valve output and process states when valve behavior cannot be neglected (e.g., stiction-induced oscillations in measured process outputs) are closed-loop effects that can be difficult to predict a priori due to the coupled and typically nonlinear dynamics of the process-valve model. Subsequently, we discuss the implications of this closed-loop perspective on the effects of valve dynamics in closed-loop systems for understanding valve behavior compensation techniques and developing new ones. We conclude that model-based feedback control designs that can account for process and valve constraints and dynamics provide a systematic method for handling the multivariable interactions in a process-valve system, where the models in such control designs can come either from first-principles or empirical modeling techniques. The analysis also demonstrates the necessity of accounting for valve behavior when designing a control system due to the potentially different consequences under various control methodologies of having different types of valve behavior in the loop. Throughout the work, a level control example and a continuous stirred tank reactor example are used to illustrate the developments. (C) 2017 Elsevier Ltd. All rights reserved.