The effect of yeast concentration on ultraviolet (UV) inactivation of five strains of Escherichia coli O157:H7 from different sources, inoculated both individually and simultaneously in orange juice, was analyzed and mathematically modeled. The presence of yeast cells in orange juice decreases the performance of UV radiation on E. coli inactivation. UV absorption coefficients in the juice increased with increasing yeast concentration, and higher UV doses were necessary to inactivate bacterial strains. UV intensities of I = 3.00 ± 0.3 mW/cm2 and exposure times (t) between 0 and 10 min were applied; radiation doses (energy, E = I × t) ranging between 0 and 2 J/cm2 were measured using a UV digital radiometer. All the tested individual strains showed higher resistance to the treatment when UV radiation was applied at 4 °C in comparison to 20 °C. UV inactivation of E. coli O157:H7 individual strain was satisfactory fitted with a first order kinetic model. A linear relationship was found between UV absorptivities and D values (radiation doses required to decrease microbial population by 90%) for each strain. The dose required to reach 5-log reduction for the most unfavorable conditions that is the most UV resistant strain, and maximum background yeast concentration was 2.19 J/cm2 at 4 °C (corresponding to 11 min of UV treatment) and 2.09 J/cm2 at 20 °C (corresponding to 10.55 min of UV treatment). When a cocktail of strains was inoculated in orange juice, the logistic equation was the best model that fits the experimental results due to the deviation from the log-linear kinetics. The UV resistance between strain cocktail and single strain were mathematically compared. Slopes of the decline curves for strain cocktail at high UV doses were lower than the slopes of the log-linear equation calculated for the individual strains, even for the most resistant one. Therefore, microbial inactivation tests using a cocktail of strains are particularly important to determine the performance of the UV inactivation treatment.