A quantitavi structure- relationship (QSRR) study has been carried out on 50 diverse phenols in gas chromatography (GC) in a dual-capillaryc olumn systemm adeo f DB-5 (SE-54b ondedp hase)a ndD B-17 (OV-I 7 bondedp hase)f used-silicac apillaryc olumnsb yusing molecular structural descriptols. Modeling of retention times of thesc compounds as a function of the theoretically derived descriptors was established by multiple linear regression (MLR), partial least squares (PLS) regression and artificial neural networks (ANN). Stepwise SPSSw as usedl or the selectiono l\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\'the variables( descriptors)t hat resultcdi n the best-finedm odels.F or predictionr etentiont imes of c:ompounds:n DB-5 and DB- l7 columns, three and four descriptors,r cspectivelyw ere usedt o develop a quantitativer elationship between the retention times and structural properties. Appropriate models with low standard errors and high correlation coeflicients were obtained. After variabless election,c ompoundsr andomlyw ere divided into two traininga nd tests etsa nd MLR and PLS methods( with leave-oneout cross validation) and ANN used fbr building ot\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\'the best models. The predictive quality of the QSRR models were tested for an external prediction set of l0 compounds randomly chosen from 50 compounds. The squared regression coefficients of prediction for the MLR, PLS andA NN modelsl br DB-5 column were 0.9645,0 .9606a nd 0.9808,r espectivelya nd also for DB-17 column were 0.9757,0 .9757 and 0.9875, respectively. Result obtained showed that non-linear model can simulate the relationship between structural descriptors and the retention times of the molecules in data scts accurately.

Keywords