A number of 2,6-bisbenzylidenecyclohexane-1-one derivatives have been synthesized and tested as HIV-1 integrase (IN) inhibitors with the aim of obtaining compounds capable to elicit antiviral activity at non-cytotoxic concentrations in cell-based assays. 3,5-Bis(3,4,5-trihydroxybenzylidene)-4-oxocyclohexaneacetic acid (20d) resulted one of the most potent and selective derivatives in acutely infected MT-4 cells (EC(50) and CC(50) values of 2 and 40 microM, respectively). In enzyme assays with recombinant HIV-1 integrase (rIN), this compound proved able to inhibit both 3'-processing and disintegration with IC(50) values of 0.2 and 0.5 microM, respectively. In order to develop a model capable to predict the anti HIV-IN activity and useful to design novel derivatives, we performed a comparative molecular field analysis (CoMFA) like 3-D-QSAR. In our model the ligands were described quantitatively in the GRID program, and the model was optimized by selecting only the most informative variables in the GOLPE program. We found the predictive ability of the model to increase significantly when the number of variables was reduced from 20,925 to 1327. A Q(2) of 0.73 was obtained with the final model, confirming the predictive ability of the model. By studying the PLS coefficients in informative 3-D contour plots, ideas for the synthesis of new compounds could be generated.