<p>This paper introduces a differentiator-based adaptive feedback linearization control method that directly estimates and compensates for lumped system uncertainties using a time-derivative estimator, thereby removing the need for explicit uncertainty models or complex adaptive structures. Conventional nonlinear uncertainty-cancelling control methods, including adaptive and robust strategies such as neural networks, fuzzy logic systems, sliding mode control, and extended observer-based techniques, are widely used to address uncertainties in nonlinear systems. However, these approaches often entail significant computational complexity, require restrictive modeling assumptions, or suffer from drawbacks such as chattering and limited stability guarantees. The proposed controller achieves robust, high-precision tracking performance without chattering and requires minimal information about the system model. Notably, the method guarantees quasi-exponential convergence of the tracking error to zero, offering stronger stability properties than those of existing approaches. It is simple to design and tune, with only two main parameters, and is broadly applicable to both affine and nonaffine nonlinear systems.</p>