This paper presents a new approach for sensor location for state and parameter estimation for stable nonlinear systems. The unique feature of this technique is that sensor locations for state estimation and measurement locations for parameter estimation can be determined within the same framework. To compute optimal sensor locations, information derived from observability covariance matrices is combined with already existing measures, which were proposed either for state or parameter estimation, to compute the degree of observability of a nonlinear system over an operating region. The optimal sensor locations then correspond to the configuration that returns the highest value of the measure for the degree of observability of a system. The proposed method is illustrated in case studies where optimal sensor locations for state and parameter estimation for a binary distillation column and a fixed-bed reactor are computed. The results obtained from the presented approach are compared with a technique based upon a linearized system.