Four isolates of Fusarium oxysporum f.sp. vasinfectum (FOV) were tested for levels of pathogenicity on 45-day-old greenhouse grown seedlings of 20 cotton genotypes. Isolates differed significantly (p=0.04) in their pathogenicity on the genotypes. Similarly, differences among genotypes were very highly significant (p=0.0000) when they were tested against the isolates. Isolate x genotype interaction was a highly significant (p=0.01) source of variation in wilt incidence suggesting that the genotypes responded differently to the different isolates. These results imply that the pathogenicity of the tested isolates is a mixture of both aggressiveness and virulence and there are significant differences among isolates in both types of pathogenicity. Similarly, resistance of the tested genotypes is also a mixture of both vertical resistance (VR) and horizontal resistance (HR) and the genotypes significantly differ in both types of resistance. Assessment of the relative contribution of each source of variation to the explained (model) variation in wilt incidence revealed that isolate aggressiveness accounted for 0.88% of the explained variation, HR of the genotypes accounted for 89.28%, and virulence of the isolates or VR of the genotypes accounted for 8.95%. The GLC analysis of amino acid composition of cottonseeds revealed the presence of 17 amino acids but the occurrence of each in the seeds varied with the genotype. Lysine and glycine were negatively (r = - 0.385) and positively (r = 0.418) correlated (p < 0.10) with VR of the genotypes to S₃ and S₄, respectively. None of the other amino acids was significantly correlated with VR of the genotypes to any isolate. Data for VR of the genotypes to each of the tested isolates (dependent variables) and concentrations of the amino acids (independent variables or predictors) were entered into a computerized stepwise multiple regression analysis. Using the predictors supplied by stepwise regression, 4 models were constructed to predict VR of the genotypes to FOV isolates. One of the generated models proved to be effective in prediction VR of the genotypes to S₂. This model showed that differences in VR to S₂ among the genotypes were due largely to the amino acids alanine, histedine, lysine, and cystine, which accounted for 61.71% of the total variation in VR. Concentrations of proline, histedine, and lysine accounted for 36.36% of the total variation in HR. The findings of the present study suggest that the variations in amino acids may, at least in part, account for the differences in the VR or HR of the different cotton genotypes to FOV.