The rice milling process produces rice husk as a by-product. It is one of the most important agricultural leftovers in terms of volume. The data of the sorption isotherm of Hg (II) (CV) sorption onto rice husk ash, which was plotted using linearized plots of isothermal models were reanalyzed using isothermal models using nonlinear regression. As the datapoints were small, nineteen isotherm models with parameters of only up to three were utilized to prevent overfitting. The models were Henry, Langmuir, Freundlich, Temkin, Dubinin-Radushkevich, Jovanovic, Redlich-Peterson, Sips, Toth, Hill, Khan, BET, Vieth-Sladek, Radke-Prausnitz, Brouers–Sotolongo, Fritz-Schlunder III, Unilan, Fowler-Guggenheim and Moreau. Statistical analysis based on error function analyses such as root-mean-square error (RMSE), adjusted coefficient of determination (adjR2), accuracy factor (AF), bias factor (BF), Bayesian Information Criterion (BIC), corrected AICc (Akaike Information Criterion), and Hannan-Quinn Criterion (HQC) showed that Freundlich model was the best model. The value of the maximum monolayer adsorption capacity for Hg binding to rice husk ash according to the Langmuir’s parameter qmL was 3.998 mg g-1 (95% Confidence interval from 2.473 to 5.523), while bL (L mg-1), the Langmuir model constants was 0.067 L mg-1 (95% C.I. from 0.001 to 0.134). The Freundlich model is unable to forecast the maximal adsorption capacity. The Halsey rearrangement of the Freundlich equation gave the estimated maximum absorption of 3.39 mg g-1, which is very close to the experimental value. The nonlinear regression method provides parameter values within the 95% confidence interval, facilitating improved comparability with prior research.