Samples of grain from New Zealand grown maize (Zea mays L.), collected from yield comparison strip trials at harvest during 1992, 1993, and 1994, were analysed for contamination by the Fusarium mycotoxins nivalenol (NIV), deoxynivalenol (DON), and zearalenone (ZEN). The samples collected in 1992 were also assessed for the incidence of infection by Fusarium spp. The samples mostly represented commercially grown hybrids, and 12 of the hybrids were sampled more extensively and were collected in each year of the study. The incidence of mycotoxin contamination was high in each year, especially by the two trichothecenes, NIV and DON. The levels found in 1992 were generally higher than in 1993 and 1994. Much of this was because of adverse weather conditions during the 1991/92 season which led to delayed harvest. In each year samples harvested after early June tended to have higher contamination levels. It was also observed in the 1992 season that Fusarium infection levels were higher in the later harvested samples. Some hybrids were more susceptible to accumulation of mycotoxin contaminants, and from the 12 extensively sampled hybrids, four could be identified which overall had higher toxin levels. These hybrids had a marked impact on the overall contamination levels found in each season studied. The findings of the study indicate that only hybrids less susceptible to mycotoxin contamination should be used in New Zealand and that cultural practices should be changed to ensure early harvesting of all crops.