Straw is a valuable by-product of crop production which can be used for various purposes (livestock feed and bedding, bioenergy). However, it should primarily be retained on farmlands to prevent soil organic matter (SOM) losses. Straw retained on the field is usually incorporated into the soil when conventional (with ploughing) and reduced tillage systems are used or left on the soil surface (mulching) when a no-tillage system is practiced. The aim of this study was to determine how different straw management practices (straw removal, straw incorporation and straw mulching) affected the incidence of Cephalosporium gramineum on two winter wheat cultivars, the grain yield of these cultivars and selected soil properties based on a long-term micro-plot experiment. Cephalosporium stripe disease was absent or occurred at very low levels (0–2.4%) when straw was removed or incorporated every second year. The disease was most severe, 24–33% tillers infected, in the SM(N) treatment with yearly straw mulching and cv. Bogatka was more tolerant to C. gramineum infection than cv. Bamberka. Importantly, yearly straw incorporation into the soil in contrast to straw mulching resulted in low disease levels (5–8% tillers infected) in both cultivars. Only in the case of cv. Bamberka was the grain yield significantly reduced in the SM(N) treatment compared to other treatments. The soil in this experiment contained the lowest level of soil SOM, which amounted to 21.0 g ∙ kg^–1 soil dry matter (DM), when each year wheat straw was removed (SR). Straw incorporation every second year resulted in 24.2 g of SOM ∙ kg^–1 soil and the largest amounts of SOM (26.0–26.1 g ∙ kg^–1 soil) were found with yearly straw incorporation into the soil. Yearly straw mulching was inferior in this respect and the soil in this treatment contained 23.8 g of SOM ∙ kg^–1 soil DM.