Charged aerosol detection (CAD) is an universal technique in liquid chromatography that is increasingly used for the quality control of drugs. Consequently, it has found its way into compendial monographs promoted by its simple and robust application. However, the response of CAD is inherently nonlinear due to its principle of function. Thus, easy and rapid linearization procedures, in particular regarding compendial applications, are highly desirable. One effective approach to linearize the detector's signal makes use of the built-in power function value (PFV) setting of the instrument. The PFV is basically a multiplication factor to the power law exponent of the equation describing the CAD's response, thereby altering the detector's signal output to optimize the quasi-linear range of the response curve. The experimental optimization of the PFV for a series of analytes is a time-consuming process, limiting the practicability of this approach. Here, two independent approaches for the determination of the optimal PFV based on an empirical model and a mathematical transformation in each case, are evaluated. Both approaches can be utilized to predict the optimal PFV for each analyte solely based on the experimental results of a series of calibration standards obtained at a single PFV. The approaches were applied to the HPLC-UV-CAD impurity analysis of the drug gabapentin to improve the observed nonlinear response of the impurities in the range of interest. The predicted optimal PFV of both approaches were in good agreement with the experimentally obtained optimal PFV of the analytes. As a result, the accuracy of the method was significantly improved when using the optimal PFV (90 – 105% versus 81 – 115% recovery rate for quantitation by either single-point calibration or linear regression) for the majority of the analytes. The final method with a PFV adjusted to 1.30 was validated with respect to ICH guideline Q2(R1).