This paper investigates the state-constrained control problem for a class of strict-feedback nonlinear systems. Existing constrained control methods, such as those with prescribed performance, typically require pre-setting constraint boundaries. This may lead to control redundancy if the set boundaries are far wider than the actual dynamic range of the system. To overcome this issue, this paper proposes a novel switching control idea of assess first, then constrain . Firstly, with the assumption that the nonlinear terms of the system and their partial derivatives are bounded, the state trajectory upper bound of the system with an unconstrained controller is analytically derived by integrating backstepping and Lyapunov stability theory. This upper bound characterizes the potential maximum amplitude of the state under unconstrained control. Based on this, a switching control strategy is then designed. a constrained controller is activated only when the artificially preset constraint boundary is smaller than this upper bound to prevent state violation. Otherwise, an unconstrained controller is employed to maintain the same closed-loop performance while avoiding the control redundancy and complexity introduced by additional terms. The proposed control strategy not only satisfies the state constraints but also achieves the desired tracking control with less control effort than the constrained control method. Simulation results confirm its effectiveness and feasibility.