Isolates of Marek's disease virus (MDV) representing three pathotypes of differing virulence were compared for relative immunosuppressive properties in genetically susceptible P2a-strain and genetically resistant N2a-strain chickens. Criteria of immunosuppression were 1) persistence of early cytolytic infection (i.e., a delay or failure to enter latency) in lymphoid organs, 2) atrophy of the bursa of Fabricius and thymus as measured by organ weight proportional to body weight at 8 and 14 days postinfection (DPI), and 3) histopathologic evidence of necrosis and atrophy in lymphoid organs. No significant differences in infection level were observed among the pathotypes during the early (4-5 DPI) period of infection. However, the extent of persistent cytolytic infection at 7-8 DPI, based on numbers of tissues positive and mean scores in immunofluorescence tests, was greater (P < 0.05) for three isolates (RK1, 584A, 648A) in the highest virulence pathotype (very virulent-plus MDV [vv + MDV]) than for two isolates (JM16, GA5) in a lower virulence (virulent MDV [vMDV]) pathotype. Results from two isolates (RB1B, Md5) classified in the intermediate very virulent pathotype (very virulent MDV [vvMDV]) fell between those from the other two pathotypes. Similarly, there was a stepwise effect of viral pathotype in which the vv + MDV isolates caused the most severe damage to lymphoid organs in terms of atrophy (relative organ weights) and histopathologic changes. Organs from chickens infected with vv + MDVs showed little recovery between 8 and 14 DPI. The vMDV isolates caused the least severe damage, and lymphoid organs showed a significant return toward normal by 14 DPI; vvMDV isolates induced intermediate degrees of atrophy and recovery. The same pattern of relationship between virulence pathotype and degree of bursal and thymic atrophy was also observed in genetically resistant N2a chickens. These results suggest that the degree of immunosuppression is linked to virulence and that a simple measure of atrophic changes (relative organ weights) in the bursa of Fabricius and thymus might be useful in determining the pathotype classification of new MDV isolates. The basis for differences in immunosuppressive potential of MDV isolates needs further clarification.