Summary1 Mycorrhizal associations of tropical, epiphytic orchids have been studied in vitro , but the importance of the relationship has not been demonstrated in the field. We investigated the effect of fungicide treatment on seedlings and juveniles of the lithophytic orchid Lepanthes rupestris . As mycorrhizas, endophytes and pathogens coexist in orchids, fungicide treatment might have either a positive or negative effect on growth and survival of plants.2 The experimental design included benomyl-treated, propiconazole-treated, and control plots. Each treatment consisted of about 140 2 × 2 cm plots divided among eight rocks, with one seedling or juvenile (occasionally two) per plot. Fungicides were applied every 2 weeks and plants were censused and measured every 3 months for 8-9 months. 3 Propiconazole-treated plants had higher mortality, significantly fewer leaves, significantly fewer fungal colonies in their roots, and significantly fewer degraded pelotons than did control plants. Benomyl-treated plants had significantly lower mortality than controls and significantly fewer fungi in leaves than control plants, and did not differ in number of degraded pelotons. There was little or no recruitment in any treatment. 4 Few potentially mycorrhizal fungi were isolated. Guignardia , Colletotrichum and Xylaria , which are generally endophytes, pathogens and/or saprotrophs, were more common. There was evidence of antagonism between Colletotrichum and Xylaria and between Colletotrichum and Guignardia . 5 Results suggest that fungi have both positive and negative effects on growth and survival of L. rupestris plants. However, conclusions about the importance of mycorrhizas are complicated by the simultaneous occurrence of a range of orchid-fungus interactions.