Samples of bovine serum from uninfected and African trypanosomes-infected animals were tested before and after gamma-irradiation, using three sandwich enzyme-linked immunosorbent assays (ELISA). Each test system utilized a different monoclonal antibody, reputedly allowing the specific detection of conserved-invariant cytoplasmic antigens of trypanonosomes, T. congolense, T. vivax, and T. brucei, respectively. Results have identified two groups of samples. The first contained samples where there were unequivocal ELISA results indicating positivity and negativity, for non-irradiated samples. In this group, irradiation had no effect on the diagnostic sensitivity of the assays. All samples shown to be positive before irradiation remained positive and those shown to be negative, remained negative. There was, however, a statistically significant reduction in signal in each of the ELISAs following irradiation. The second group contained samples identified before irradiation as flanking the diagnostic negative/positive threshold of OD > or =0.05. These showed a negative bias after irradiation of the order of OD -0.01, which was shown to be statistically significant by paired t-statistics. Without correction of the given diagnostic negative/positive threshold, bovine sera with OD values around the threshold were expected to deliver more false negative test results upon irradiation. This was confirmed when serological data were compared with parasitological findings; where three times more false negative test results were found from irradiated serum samples. Consequently, for this group of irradiated bovine samples tested by ELISA, the re-adjustment of the diagnostic negative/positive threshold of the ELISAs using defined irradiated serum samples is recommended; otherwise, the frequency of false negative results might be increased.