In this thesis, we develop a passivity-based adaptive control framework for controlling nonlinear processes with uncertainty. The development of the method is motivated by the question whether we can control reaction systems without the knowledge of reaction kinetics. The proposed adaptive control framework incorporates the measurements’ derivative information in order to estimate the uncertainty involved in output dynamics. The output dynamics is assumed to take a special control-affine structure, and by using the output’s derivative information we can avoid using internal state dynamics, which is not usually available. Passivity theory is applied for control and estimation designs and overall closed-loop stability is achieved. By extending the passivity-based control to systems with relative degree higher than one through backstepping, we can obtain cascade feedback schemes with PID controllers that overall control convergence is guaranteed. The proposed framework allows us to control reaction systems without knowing the reaction kinetics, and estimate unmeasured compositions by utilizing the available partially linear structure of internal dynamics. A reactor temperature control problem that usually has high relative degree is used to illustrate the application of passivity-based backstepping control, and results from industrial trials are presented.