An electrically conductive polymer composite (CPC) based on polycarbonate filled with 1.5 wt.% carbon nanotubes (CNTs) was investigated regarding its solvent selectivity when used as a sensor material for liquid detection. Based on the electrical response characteristics of the CPC when immersed in different solvents, "good" and "bad" solvents out of 59 test liquids were detected and the Hansen solubility parameters of the CPC were calculated using the "Hansen software" to be delta(D) = 18.4 +/- 0.2 MPa0.5, delta(P) = 10.9 +/- 0.8 MPa0.5, and delta(H) = 4.1 +/- 0.5 MPa0.5. Based on the CPC's and solvents' Hansen solubility parameters the affinity between CPC and solvent represented by the distance in Hansen space was determined. As a second parameter the molar volume of the solvent molecules was used to describe the selectivity of the CPC by means of a quarter circle like area clearly separating "good" and "bad" solvents in a so-called solvent map. Whereas the Hansen solubility parameters are based on thermodynamics, the influence of the solvents' molar volume on the CPC's selectivity can be explained by diffusion processes. When increasing the molar volume of solvents having a similar chemical structure, a limiting value of the molar volume was found above which no solvent diffusion into the CPC was observed. Using this large number of solvents it could be clearly shown that the electrical response kinetics upon immersion into "good" solvents cannot be correlated with the Hansen solubility parameters of the CPC or their difference between CPC and solvent but is determined by the diffusion kinetics, which is governed by the solvents molecule size. (C) 2012 Elsevier Ltd. All rights reserved.