Recently synthesized dialkoxybenzamide derivatives, structurally analogous to Roflumilast with more selectivity against phosphodiesterase 4B, were highlighted in the present work. To find optimum chromatographic conditions for the elution of these compounds, a central composite experimental design was carried out by varying the stationary phase type as a categorical factor and mobile phase composition including the percentage of acetonitrile and the pH of the buffered water as continuous factors; the obtained retention times were utilized in the quantitative structure retention relationship (QSRR) studies. Furthermore, quantitative structure activity relationship (QSAR) studies for these promising compounds were performed. QSAR and QSRR models were built by different techniques namely multiple linear regression (MLR), principal component regression (PCR) for linear modelling and Principal Component-Artificial Neural Networks (PC-ANN) for nonlinear modelling. Internal validation (leave many out method), and external validation were used to evaluate the performance of the generated models. Depending on the calculated statistical parameters, PC-ANN QSAR model showed the best predictive power for the biological activities of the test set compounds. Whereas MLR technique was more suitable to build QSRR model, this model can help in understanding how the chemical structure and the lipophilicity of the compounds can affect the retention time and chromatographic behavior.