The content of soil mineral N (N min.) in early spring and after the harvest of crops (beginning of September) was studied in two sites at Lipa and Svitavy in the years 1991 to 1993 in long-term stationary treatment trials. In both sites, at Lipa since 1974 and Svitavy since 1980, 12 treatments of fertilization have been included in trials. Trials were arranged in eight-plot crop rotation with 50% cereals, 25% forage and 25% root crops. The rotation of crops was identical in both sites and the following crops were cultivated in the years under observations: 1991 - red clover, 1992 - winter wheat, 1993 potatoes followed by winter wheat. Nitrate and ammonium N were determined from freshly taken soil samples in extract of 1% K2SO4 by ion selective electrodes. High content of mineral N was found in 1998 in the spring in the clover stand (20.7 to 40.1 ppm N), but also in autumn after incorporation of clover (20.7 to 30.1 ppm). Most of determined mineral N was in ammonium form. Marked increase in mineral N was found in autumn in the year 1993 in treatments fertilized by gradated doses of NPK in combination with manure (applied in autumn of 1992). Gradated fertilization of P and K did not affect more significantly the content of mineral N. Positive correlation between mineral N content and wheat yield was found. This relationship was not confirmed in potatoes. Positive correlation between mineral N content was determined, in the spring and uptake of N by wheat and in potatoes in the autumn.