A method was developed for monitoring the growth of a range of bacteria and fungi in the Bactometer 32 impedimeter. 195 yeast strains, 74 strains of mould and 20 strains of bacteria were screened for sensitivity to 1 mug ml[-1] of T-2 toxin. Growth inhibition was assessed impedimetrically. Twelve bacteria, 20 moulds and 38 yeast strains were tested against roridin A, verrucarin A, deoxynivalenol, and diacetoxyscirpenol each at 1 mum ml[-1] . Deoxynivalenol was found to be only slightly toxic. Roridin A and verrucarin A were markedly more toxic than the others to the fungi, but not to the bacteria. Four microorganisms (Bacillus subtilis, Candida albicans. Hansenula fabianii and Pichia burtonii)were selected for further study. No obvious pattern of response could be discerned for the effects of different carbohydrates on microbial sensitivity to T-2 toxin. The effects of T-2 toxin on B. subtilis and C. albicans were greater if chloroform rather than dichloromethane and methanol (95:5, v/v) was used as the toxin carrying solvent, the reverse was true for H. fabianii and P. burtonii. Dose-response curves were constructed, based on impedimetric responses, and no-effect levels were determined for each species. For B. subtilis the no-effect level was 0.35 mum ml[-1], for C. albicans it was 0.40 mum ml[-1], for H. fabianii it was 0.012 mum ml[-1] and for P. burtonii it was 0.018 mum ml[-1]. H. fabianii was selected for further studies. This strain was found to be auxotrophic for proline. The effect of T-2 toxin and verrucarin A on the uptake of radiolabelled proline in H. fabianii cultures was studied. Dose-response curves were constructed for each toxin. It was found that both toxins reduced proline uptake in a dose dependent manner, the no-effect levels were 0.025 mum ml[-1] for T-2 toxin and 0.0125 mum ml[-1] for verrucarin A.