Root-knot nematode (Meloidogyne javanica) management should be partly based on the knowledge of the threshold density, and this value is likely to vary depending on the resistance level of the tomato cultivars. The damage functions based on four initial population densities (Pi) (0, 1,000, 3,000 and 5,000 egg kg^-1 of soil) of root-knot nematode were determined in four tomato cultivars. The experiment was performed in completely randomized design with four replications. The results showed that yield responses to Pi were fitted properly by a log-logistic function with three parameters. The most susceptible cultivar was ‘Rutgers’ based on EP₅₀ and EP₁₀ (effective population of nematodes, reducing 50% or 10% of maximum yield or shoot fresh weight respectively) and three others were relatively resistant. EP₁₀ is more applicable than EP₅₀ because 50% yield reduction is unacceptable in most situations. EP₁₀ for yield of ‘Rutgers’, ‘Efialto’, ‘Falat 111’, ‘Gina VF’ was 500, 3,021, 2,998, and 3,000 egg kg^-1 of soil, respectively. The correlation coefficients among gall index, egg mass and reproductive factors were positively related. Reproduction factor and root gall indices were greater in ‘Rutgers’ than in the other cultivars (P≤0.05). For ‘Gina VF’ as a relatively resistant cultivar it seemed that increasing of Pi up to 5,000 or more egg kg^-1 soil might break its resistance.