Sixteen tomato [Solanum lycopersicum L. (syn. Lycopersicon esculentum Mill.)] genotypes (inbred lines or hybrids) were tested against five Phytophthora infestans (Mont.) deBary isolates to characterize race specificity of late blight resistance transferred to tomato from Solanum pimpinellifolium L. [syn. Lycopersicon pimpinellifolium (L.) Mill.] accession L3708. The effects of plant genotype, isolate, genotype × isolate, and isolate × replication interactions were highly significant (P = 0.001). Set of four sister lines fixed for late blight resistance (CU-R lines) exhibited full and equal resistance to the five pathogen isolates tested. In contrast, the heterozygous F1 hybrids, created by crossing the resistant CU-R lines with a susceptible parent, were resistant to US-11; partially resistant to US-17, NC-1, and DR4B; and susceptible to US-7. Differential responses were also observed across pathogen isolates on a set of resistant sister lines (CLN-R lines), which also were bred from L3708. The CLN-R lines were resistant to the DR4B, NC-1, and US-11 isolates, but showed significant disease-affected areas and sporangium numbers following inoculation with either US-7 or US-17. Restriction fragment length polymorphism (RFLP) analysis confirms that both CU-R and CLN-R are homozygous for the Ph-3 gene derived from L3708. Since progeny tests also confirmed that the CLN-R lines are fixed for their level of resistance, these results suggest that late blight resistance in the CU-R lines is not controlled by Ph-3 alone, and that at least one additional gene conferring late blight resistance is missing from the CLN-R lines. Results of genetic tests of the (CU-R × CLN-R) F1 and a (CU-R × CLN-R) F2 population with the pathogen isolate US-17 strongly support a model in which resistance of the CU-R lines requires genes in addition to Ph-3. The implications of this information in breeding for late blight resistance and using of the resulting resistant lines or hybrids are discussed.