Introduction: Solubility is an important thermodynamic property due to its role in product development, as well as the understanding of biological processes. This research aims to evaluate the preferential solvation parameter (δx1,3) of the triclocarban (TCC) solubility in N-methyl-2-pyrrolidone + water cosolvent mixtures and to assess some correlational and predictive mathematical models of concern to the pharmaceutical industry. Calculations: δx1,3 was determined from experimental data following the Inverse Kirkwood-Buff Integrals model (IKBI). The mathematical models were developed using Python, and functions for each model were fitted by non-linear least squares using the libraries scipy.optimize. curve_fit, and sklearn.model_selection. Results: According to the δx1,3 heat, TCC has preferential solvation by water in water-rich mixtures, and preferential solvation by N-methyl-2-pyrrolidone in intermediate and N-methyl-2-pyrrolidone-rich mixtures. The models Yalkowsky–Roseman–van’t Hoff, Wilson, Modified Wilson, NRTL, van’t Hoff, Apelblat, and Buchowski–Ksiazaczak 𝜆ℎ were assessed, finding good correlations with all. Conclusions: The TCC solubility increase in N-methyl 2-pyrrolidone + water cosolvent mixtures with increasing N-methyl-2-pyrrolidone (NMP) concentration may be related to the rise in the local mole fraction (x1,3 L) of NMP in the TCC solvation sphere. Regarding the mathematical models, the Yalkowsky–Roseman–van’t Hoff model can be considered the most versatile due to its capability estimate solubility data as a function of both temperature and cosolvent composition, given a limited range of experimental data.