The salinity tolerances of 21 accessions belonging to four wild tomato species [Lycopersicon pimpinellifolium (Jusl.) Mill., L. peruvianum (Corr.) D'Arcy, L. hirsutum (L.) Mill., and L. pennellii Humb. Bonpl.) were evaluated using their vegetative yield-salinity response curves at the adult stage, determined by a piecewise-linear response model. The slope (yield decrease per unit salinity increase), salinity response threshold, maximum electrical conductivity without yield reduction (ECo), and salinity level for which yield would be zero (ECo) were determined by a nonlinear least-squares inversion method from curves based on the response of leaf and stem dry weights to substrate EC. The genotype PE-2 (L. pimpinellifolium) had the highest salt tolerance, followed by PE-45 (L. pennellii), PE-34, PE-43 (L. hirsutum), and PE-16 (L. peruvianum). The model also was tested replacing substrate salinity levels with leaf Cl- or Na+ concentrations. Concentrations of both ions for which vegetative yields were zero (Clo and Nao) were determined from the response curves. In general, the most tolerant genotypes were those with the highest Clo and Nao values, suggesting that the dominant salt-tolerance mechanism is ion accumulation, but there were cases in which salt tolerance was not related to Clo and Nao.