BACKGROUND AND AIMS: Litter is a key factor in structuring plant populations, through positive or negative interactions. The litter layer forms a mechanical barrier that is often strongly selective against individuals lacking hypocotyle plasticity. Litter composition also interacts with plant growth by providing beneficial nutrients or, inversely, by allowing harmful allelopathic leaching. As conspicuous litter fall accumulation is often observed under deciduous forests, interactions between tree litter and understorey plant populations are worthy of study. METHODS: In a 1-year ex-situ experiment, the effects of tree litter on the growth of Anemone nemorosa, a small perennial forest geophyte, were investigated. Three 'litter quantity' treatments were defined, representative of forest floor litter (199, 356·5 and 514 g m(-2)), which were crossed with five 'litter composition' treatments (Quercus petraea, Fagus sylvatica, Carpinus betulus, Q. petraea + F. sylvatica and Q. petraea + C. betulus), plus a no-litter control. Path analysis was then used to investigate the pathways linking litter characteristics and components of adult plant growth. KEY RESULTS: As expected, the heavier the litter, the longer the petiole; rhizome growth, however, was not depreciated by the litter-induced petiole lengthening. Both rhizome mass increment and number of initiated buds marginally increased with the amount of litter. Rhizome mass increment was in fact determined primarily by leaf area and leaf life span, neither of which was unequivocally correlated with any litter characteristics. However, the presence of litter significantly increased leafing success: following a late frost event, control rhizomes growing in the absence of litter experienced higher leaf mortality before leaf unfolding. CONCLUSIONS: The study questions the role of litter as a physical or chemical barrier to ground vegetation; to better understand this role, there is a need for ex-situ, longer-term experiments coupled with in-situ observations in the forest.