This study aims to determine changes in the structure of cryptogamic vegetation of poor psammophilous grassland along a pollution gradient near a zinc smelter to evaluate the potential of species assemblages as bioindicators of soil condition. Lichens and bryophytes were examined in study plots along six transects in four distance zones, and the physicochemical properties of corresponding soil samples were analysed. Four different responses of species to substrate contamination were identified, with a distinct group of species resistant to and favoured by metal contamination. Although species richness decreases as one approaches the smelter, the gradual replacement of certain sensitive species by resistant ones was observed along the pollution gradient. The results enabled us to develop a useful tool to diagnose strongly polluted sites. Two different cryptogamic assemblages of well-recognised key species characteristic for strongly polluted and lightly polluted sites were distinguished. We conclude that cryptogamic community structure clearly corresponds to the degree of soil contamination, thus demonstrating high bioindicative value. The study confirmed the high relevance of the community approach in metal pollution biomonitoring.