Accessions of Lycopersicon esculentum var. cerasiforme (Dun.) A. Gray (cer) and L. pimpinellifolium (Jusl.) Mill. (pimp), sustained significantly less damage to fruit by beet armyworm [Spodoptera exigua (Hiibner)] than standard cultivars and breeding lines of L. esculentum Mill. (esc) under natural infestations in 1990 and 1991 in southern California. The dwarf vine cherry cultivar Tiny Tim also sustained less damage than the standards. Accessions of esc with various monogenic mutations sustained at least as much beet armyworm damage as did standard cultivars. The percentage of fruit damaged was significantly correlated with vine weight, weight per fruit, number of fruit, and the fruit-foliage weight ratio (Pearson's coefficients, respectively: -0.533, 0.450, -0.483, 0.390, n = 37). In laboratory assays, survival of beet armyworm was significantly lower (5% of susceptible& growth rates were significantly lower, and development time was significantly longer on the fruit of resistant `Tiny Tim' and LA 1320 cer than the fruit of 11 other test lines. There were no substantial differences in beet armyworm survival on the foliage of the test lines. In the field trials, there were also significant differences among the test lines in damage by Liriomyza species and hemipteran pests. Lines with genes for increased densities of nonglandular leaf trichomes (especially LA 1663) were generally less damaged by Liriomyza than other lines. Damage by hemipterans was correlated with vine and fruit size, fruit count, and fruit-foliage weight ratio in 1991, but high intraseason variability prevented clear identification of test lines resistant to these pests.