A total of 84, 78, 80, 80, 78, and 45 wheat samples for feed use were collected randomly after the 1987, 1989, 1990, 1991, 1992, and 1993 crops, respectively, from farms in an area of southwest Germany. The sum of precipitation from May to September varied during years, with markedly higher precipitation in 1987, compared to 1989-1993. Deoxynivalenol, 3- and 15-acetyldeoxynivalenol, nivalenol, HT-2 toxin, T-2 toxin, diacetoxyscirpenol, and fusarenon-X were determined by gas chromatography, combined with mass selective detection (GC-MS), zearalenone, alpha- and beta-zearalenol by GC-MS or HPLC. Deoxynivalenol was the major toxin, with incidences of 68-95% and mean contents at 152-1,692 micrograms/kg. In contrast, incidences of zearalenone, 3-acetyldeoxynivalenol, nivalenol, HT-2 toxin, and T-2 toxin were at 11-80, 17-60, 25-64, 0-8, and 0-41%, respectively, with mean contents between 3 and 209 micrograms/kg. alpha-zearalenol and/or beta-zearalenol were detected in five samples at contents < or = 71 micrograms/kg; diacetoxyscirpenol was not detected in any sample. 15-acetyldeoxynivalenol and fusarenon-X were assayed in samples from 1987 and 1991-1993. 15-acetyldeoxynivalenol was detected in 3-15% of samples at mean contents of 5-84 micrograms/kg; fusarenon-X was not detected. Over the years, incidences and levels of toxins remained constant, decreased or increased. The correlation between the occurrence of toxins and precipitation is discussed.