The use of the multi-mycotoxin screen based on dialysis to analyze foods and feeds for mycotoxins, is well documented. This study investigated the possibility of incorporating FB I into the screen. Maize meal (25g) was spiked with AFB I , CPA, FB1, ST and ZEA and extraction was done using acetonitrile/4% potassium chloride (90:10 v/v). The recoveriesof the mycotoxins were 77.4, 61.5, 97.4, 79.8 and 98% respectively on analysis by HPLC. Fumonisin B1 could not be completely incorporated into the screen due to its reaction with sodium hydrogen carbonate, which is a component in the method. Thus, FB I was determined in a separate portion of the extract. The high cost of FBI standards which are often of inferior purity necessitated that FB I standards be locally produced in the laboratory using Fusarium moniliforme MRC 826, a good producer of FB 1 . In this study, production of FB I was carried out using a stirred jar fermenter and patty cultures. The yields were 160mg/1 and 6mg/g of FB I for the two methods respectively. Methyl esterification of tricaballylic acid moieties of FB I was done for effective clean-up. This was achieved by derivatizing FBI, with diazomethane. It was found that other functional groups besides the tricaballylic acid moieties of FB I were undesirably methylated as well, which made cleanup by this method difficult as shown by electrospray mass spectrometric analysis. Attempts to de-methylate FBI methyl esters with esterase was not successful. Analysis of human faecal samples was carried out with the view of developing a short term marker for assessing human exposure to FB I . Faeces from rural (20) and urban (23) volunteers were analyzed by high performance liquid chromatography. The results showed that 35% of the rural samples and 9% of the urban volunteers had detectable amounts of FB I ranging from 0.600 to 19.56 mg/kg. There was a significant difference (p = 0.04)between the two population groups. A study was carried out to assess the occurrence of FBI in a rural area of Tugela valley in Kranskop magisterial district of KwaZulu Natal. A questionnaire was administered to gather information on the family health and nutrition. Raw (stored) and processed foods and faeces, were collected for analysis of FB1. A similar control study was carried out in the urban area of Durban Metro. Homes were mapped out using the GIS for easy follow up. Oesphageal cancer (OC) incidence from the local hospital and weather data for the study area were collected from South African Weather Bureau, Johannesburg. The questionnaire results showed that the common diseases were mainly of respiratory origin (24% and 26%) from both rural and urban groups respectively. Food analysis (by HPLC) showed that the number of maize samples with FB I were higher in the rural area (31.9%) in comparison to the urban samples (6.1%). The level ranged from 0.092-22.225 mg/kg in food and 0.513-39 mg/kg in faeces. The mean concentration of FB i in the faeces and maize samples showed a similar significant difference of 0.014 between the two groups. However, these concentrations were much lower than those of high OC area in Transkei (117 mg/kg). There was no detection of FBI in fermented food products. / Thesis (M.Med.Sci.)-University of Natal, Durban, 1998.