It was demonstrated that a mechanical harvesting regime can influence the species composition of ditch vegetation at the community level. This effect, however, was very small compared with those of other factors such as the between-site and the within-site spatial variation, and several soil and water quality parameters. Cutting in November had the largest effect, in that it caused the greatest extremes in species cover. The vegetation was composed of 136 plant species. The semi-aquatic and aquatic species were less numerous than the terrestrial ones (52 versus 84). The total number per vegetation type and site ranged from 5 to 49; that of persistent plant species, from 4 to 22. Only 16~o of the species was significantly influenced by the mechanical harvesting regime. The significant effects of the mechanical harvesting regime on plant species were related to plant-inherent factors. Mechanical harvesting repeated within a year, on one hand, opened up the vegetation, 1) freeing sites for colonization of new species, 2) improving the light climate for seedlings which had already colonized; on the other hand, it exhausted (carbohydrate) reserves of solitary species. Mechanical harvesting once a year in November had a contrasting effect in that it caused suffocation of the shore vegetation in spring by the not yet decomposed plant material harvested in autumn. The highest species richness was attained for the aquatic vegetation: on sand by cutting three times per year (in May, July and September), and on peat by cutting once a year (in November); and for the shore vegetation: on sand once a year (in May) and two times per year (in May and July), and on peat once a year (in May or November).