The partition coefficients, P, between n-octanol and water of a number of growth stimulating substances, N-hydroxyethylamide of aryloxyalkylen- and pyridine carboxylic acids were obtained from Pomona College ( C log P), and Rekker's (log P Rekker) revised fragmental constant system was used to calculate log P data sets. Both of these data sets were correlated with two different substance lipophilicity parameters, log k w and ϕ 0. Log k w was obtained by extrapolation of log retention factor ( k) to 0% organic modifier measured in reversed-phase liquid chromatography (RPLC) systems. ϕ 0 values were obtained from the slopes and intercepts of these relationships. The RPLC experiments were performed on four commercially available reversed-phase columns. Binary mixtures of methanol–water, methanol–phosphate buffer (pH 7.0), methanol–tricine buffer (pH 7.0) and acetonitrile–water were used as mobile phases for the determination of log k w values. For the methanolic eluents linear regression provided satisfactory correlations ( r>0.99) for the relationships log k vs. organic modifier content in the eluent, while for the acetonitrile-containing eluents a second-degree polynominal regression was necessary. For all four RPLC columns, by linear regression satisfactory correlations ( r>0.99) were obtained between log k w and log P data using methanolic eluents. In such eluents ϕ 0 values were shown to be the second-best lipophilicity parameters. For acetonitrile-containing eluents the use of second-degree polynominal regression was necessary and, in contrast to methanol, significant influence of the applied column on regression results was observed. For acetonitrile-containing eluents the ϕ 0-index does not provide satisfactory results for our substances. No difference in regression results between the use of buffered and non-buffered eluents was observed.