Abstract Whether production of winter wheat is stable without any fertilizer input on Greyic Phaeozem in the Czech Republic still remains unsolved. Phaeozems represent 3% of soils in Europe and they are particularly common in wet steppe regions. In this paper, the yield of winter wheat over 50 years of farmyard manure (FYM), N, P and K fertilizer application (12 treatments altogether) was analyzed in the Čáslav Crop Rotation Experiment (CCRE), which was established in 1955 in a sugar beet growing area (mean annual precipitation and temperature 555 mm and 8.9 °C, respectively). In the unfertilized control treatment, low plant available P and suitable K concentrations were recorded after 50 years. The annual yield growth (AYG) of grain was positive, even in the control treatment without any fertilizer input. The mean AYG ranged from 24.9 kg ha −1 in the control treatment to 73.6 kg ha −1 following the application of 119 kg N ha −1. In the first two decades of the experiment, there was no significant effect of treatment on grain yield. The increase in grain yield per 1 kg of applied N was 20.1 kg ha −1 in the last decade of the experiment. The mean grain yields of long-straw and short-straw varieties were 3.8 and 5.5 t ha −1, respectively. In the control, the grain yields were 4.8, 4.1 and 3.3 t ha −1 after legumes, root crops and cereals, respectively. To summarize the 50 years of results of winter wheat production in the CCRE, grain yield was most influenced by variety, followed by mineral fertilizers, the preceding crop and only moderately by FYM application. The highest grain yields were recorded in treatments with the highest N application rate. The long-term stability of winter wheat production on Greyic Phaeozem can be expected under a crop rotation system with legumes and root crops, even without any fertilizer input.