Question: Are the recruitment patterns of deliberately introduced wildflower species influenced by cutting frequencies and disturbance treatments? To what extent do these different treatments affect productivity and sward structure of an agriculturally improved grassland? Location: A mesic lowland grassland near Göttingen, Lower Saxony, Germany. Methods: Recruitment success of eight sown wildflower species was studied in a permanent grassland treated by a factorial combination of different pre-sowing cutting intervals (1, 3 or 9 wk), post-sowing cutting intervals (1,3 or 9 wk) and disturbance (control, harrowing, removal of sward). Seedling emergence and survival, biomass production and sward structure were followed over two years. Results: For most species seedling emergence was highest in the harrowing treatment. The complete sward removal did not further increase seedling emergence. Seedling survival was strongly influenced by the post-sowing cutting frequency with highest mortality in the 9 wk cutting interval compared to one and 3 wk cutting intervals. Annual dry matter yield varied between 4.4, 5.9 and 9.4 t.ha-1 in the 1,3 and 9 wk pre-sowing cutting treatment, respectively. In June, when the seeds were sown, the tiller number of the 1 wk cut plots was twice as high as for the 9 wk cut plots and five times higher than in the harrowing treatment. Conclusions: Disturbance by harrowing provided the optimal environmental cues to trigger germination, whereas seedling survival was facilitated by increased light penetration due to frequent cutting. The investigation revealed the overriding importance of frequent standing crop removal in the early phase of seedling establishment on agriculturally improved grassland.