Fusarium wilt caused by the fungus Fusarium oxysporum f sp. lycopersici is a disease of tomato (Lycopersicon esculentum Mill.) with major economic importance. The most practical form of controlling this disease is through genetic resistance from related germplasm. Single dominant genes from related species have been used to control the pathogen in the past but these are generally overcome by new races after some years of field planting resistant material. The gene 1-3 is successful to date in controlling Race 3 Fusarium wilt, but there is concern that the pathogen will overcome 1-3 and again cause serious losses to the industry. One way of preventing this could be to accumulate resistance genes from different sources into elite breeding material. Use of molecular markers would facilitate accumulation of resistance genes in elite breeding material. The aims of this study were: to resolve the nature of the underlying genetics of a new source of Fusarium wilt resistance using traditional inheritance studies and molecular marker technologies,· to develop a molecular marker map of the tomato genome for the purpose of identifying genomic regions linked to resistance, and; to assess the potential use of molecular markers as a tool in tomato breeding and introgression of resistance genes. The segregating populations used in this study were from a cross between a susceptible cultivar, Rouge de Marmande, and the resistant breeding line, 28A-1. Resistance in 28A-1 was derived from an interspecific cross involving BTN472, a resistant accession of the species L. pimpinellifolium. Fifty-one BCj, 128 F2 plants and F3 progeny from seven susceptible F2 plants were used for the phenotypic analyses. Vegetatively propagated cuttings from each plant of the parent, F1, F2, and BC1 populations were tested for their reaction to Race 1, Race 2, and Race 3 of Fusarium wilt. For Race 3, the F2 plants segregated in a ratio consistent with both a single dominant gene (3:1) and a two-gene model where a dominant and a recessive gene are acting independently (13:3). Either model could not be confirmed unequivocally in the BC1. To determine which of these two models best described the resistance to Race 3, F3 progeny from seven susceptible F2 plants were inoculated with Race 3 Fusarium wilt. Recovery of resistant F3 plants supported the two-gene model, as no resistant plants were expected from susceptible F2 plants if a single dominant gene was responsible. The single dominant gene for resistance to Race 2, l-2, was detected by its close linkage to the RFLP marker TG105. TG105 segregated in the F2 population in a ratio consistent with the expected 1:2:1 model. However, only one F2 plant (out of 116 tested) was susceptible to Race 2. This could be a reflection of the poor test conditions, as three out of ten susceptible parent plants were rated resistant to Race 2. Alternatively, I-2 segregation clearly did not correspond with segregation for disease ratings, which suggests that 28A-1 carries another genetic resistance to Race 2, non-allelic with I-2, and that it and the l-2 gene segregated in the F2 and BC1 populations. Two non-allelic genes may also control Race 1 resistance in 28A-1. Three molecular marker techniques were used for the mapping and quantitative trait locus (QTL) analysis: RFLP, RAPD, and AFLP. All three techniques were successfully applied during this study to generate DNA profiles of tomato. A comparison of the three techniques led to the conclusion that AFLP was the most efficient for generating markers, providing technical expertise was available for running denaturing polyacrylamide gels. Ninety-five F2 plants were used to generate two coupling-phase linkage maps and to identify genomic regions contributing to Race 3 resistance. The maps consisted of 157 markers (six RFLP, 71 RAPD, and 80 AFLP) covering about 615 cM QTL analysis led to the identification of three independent genomic regions implicated in Race 3 resistance. The third region, from the resistant parent, was unexpected and had a negative influence on resistance. This region was associated with markers influenced by segregation distortion, with selection occurring against these alleles. These results clearly demonstrated the utility of molecular marker and mapping techniques for understanding the inheritance of a multigenic trait. RAPD and AFLP were also used to screen two pooled DNA samples made from F2 plants at opposite extremes of the Race 3 distribution. This experiment was successful in identifying markers linked to the same three regions as identified in the QTL analysis, thus corroborating the results. Molecular markers have great potential to facilitate introgression from wild species to commercial tomato cultivars.