In this paper, the endpoint is the application of molecular topology to the search of QSAR relations into a group of dication-substituted carbazoles, furans and benzimidazoles, all showing antifungal activity against C. albicans. Mathematical and statistical methods such as linear regression and discriminant analysis, are used to goal. The obtained results clearly show a high efficiency of the formalism on the prediction and classification of antifungal activity. 83% of the compounds showing MIC<10μg/ml (active group) are correctly classified, whilst 100% overall accuracy is achieved for those compounds showing MIC>100μg/ml (inactive group).