ABSTRACT Plants have been reported as used by local populations to treat various infections for a long time, which has directed several pharmacological studies. The main aim of this work was to evaluate three plant selection criteria with better predictive power to detect extracts with antifungal action: (1) medicinal plants that are not used for indications of infection and inflammation; (2) plants with direct citations for inflammation, except for infection; (3) plants with direct citations for inflammation and infection selected quantitatively by Syndromic Importance Value (SIV). We tested the action of 23 hydroethanolic extracts of plants against the fungi Candida albicans, Cryptococcus neoformans, and Cryptococcus gattii and found no differences in the number of active extracts among the different strategies used, but activity quality varied. The extract of Anacardium occidentale presented fungicidal activity against the three analyzed fungi. At least five species - A. occidentale, Myracrodruon urundeuva, Poincianella pyramidalis, Anadenanthera colubrina var. cebil, and Mimosa oftalmocentra - presented fungistatic and fungicidal effects against all strains. Our findings indicate that selecting plants based on popular indications and quantitative prioritization techniques increases the chance of detecting potential antifungal candidates, and that the plants selected by these criteria were more effective against C. neoformans.