Six forms of fertilizers and three rotations have been examined for 120 years in the ”︁Eternal Rye trial” in Halle (central German arid region, Haplic Phaeozem on sandy loess). Rank analysis, trend analysis and a new component model have been found to be suitable statistical methods to evaluate long‐term results without true replications. In this trial it was shown that mineral fertilization maintains like farmyard manuring the yield potential although at lower content of soil organic matter (SOM). Without fertilization, yield decline was greater with potato and silage maize than winter rye (Secale cereale L.). Deficiencies in N fertilization lead immediately, and in PK supply after more than 20 years, to a significant yield decrease which could be quickly compensated by full fertilization. Rye and maize monocultures also resulted in yield decreases. The establishing of new steady states in the turnover of SOM took about 50 years after changes in fertilization. Contrary to rye monoculture, both silage maize monoculture and potato alternating with winter rye caused considerable decomposition of SOM. According to analytical pyrolysis (Py‐FIMS and Py‐GC/MS), fertilization affects the SOM composition more than rotation. Without fertilization, a higher percentage of thermically stable SOM remained in comparison to the FYM soils. The introduction of potato into the rotation enhanced the content of easily decomposable SOM.