Derivative based learning rule poses stability problem when used in adaptive plant modeling. In addition the performance of these techniques deteriorates when used for non-linear plant modeling. In this chapter, the plant modeling task is formulated as an optimization problem. A recently introduced evolutionary algorithm, cat swarm optimization (CSO), is used to develop a new population based learning rule for the model. Adaptive modeling of a benchmarked plant is carried out through simulation study. The performance of the CSO in presence of nonlinearity in the plant is also studied. The results demonstrate superior performance of the CSO compared to that achieved by genetic algorithm (GA) and particle swarm optimization (PSO) based approaches for adaptive modeling.