Question Along river floodplains lower distribution limits of plant species seem largely determined by their tolerance to rarely occurring floods in the growing season. Such distribution patterns remain fixed for many years suggesting additional effects of winter floods at lower positions. Our objective was to investigate the direct and indirect effects of winter floods on colonization of floodplains in a series of field experiments.Location: River Rhine, The Netherlands.Methods: We measured the direct effects of winter floods on seedling survival and seed removal and survival at low and high floodplain elevation. Indirect effects of winter flooding through changes in the soil were investigated by measuring seedling emergence on soil transplants that were exchanged between high and low floodplain elevation. To investigate indirect effects of floods on the germination environment through changes in the vegetation structure, we measured the effects of vegetation removal on recruitment of sown species. Results: Recruitment was seed limited at both floodplain elevations. An additional effect of vegetation removal on seedling emergence was also observed. Soil types from both zones did not differently affect seedling emergence. Seeds were not removed from the soil surface by a single winter flood. Moreover, seeds remained viable in the soil for at least two years, while the experimental plots were flooded several times during the experimental period. During one of those floods a thick sand layer was deposited at the low zone and subsequently no seedlings were observed anymore.Conclusions: Colonization of low floodplain zones in years between subsequent summer floods is prevented by seed limitation while the direct effects of winter floods are limited except for irregularly occurring sand depositions.Nomenclature: van der Meijden (1996).