A reliable set of data on the bioconcentration factors (KB) of a diverse range of compounds in fish was selected from the literature. Using the structures of these compounds, the following molecular parameters were calculated: molecular weight (MW), solvent accessible molecular surface area (SASA), solvent accessible molecular volume (SAV), molar refraction (MR), largest principal moment of inertia (LPMI) and several molecular connectivity indices of the Randić type (1 chi, 2 chi, 3 chi, 1 chi vr, 3 chi c). The relationships between these parameters and log KB were evaluated for all compounds and the following subgroups: chlorinated hydrocarbons (CHC), polyaromatic hydrocarbons (PAH), and CHC and PAH combined. These relationships indicated that SASA, SAV, and MR were good predictors of log KB for the CHC and PAH combined or alone and the other parameters were less satisfactory with these groups. In addition with the CHC, the log of these parameters displayed an improved correlation with log KB due to apparent nonlinearity in the log to linear relationship. Thus, with these groups of compounds, calculated values of SASA, SAV, and MR provide a satisfactory means of estimating log KB without measured data.