Sensory ganglia latently infected with herpes simplex virus (HSV) were transplanted beneath the renal capsule of syngeneic recipients, and the latent infection remaining was investigated. HSV latency-associated transcript (LAT) expression and reactivation of HSV after explant of transplanted dorsal root ganglia were monitored as markers of latency. Two to four weeks after transplantation, both indicated evidence of HSV latency in transplants. At those times, infectious virus was not detected in direct ganglion homogenates. In addition, viral antigen and infected cell polypeptide 4 RNA were not detected. Taken together, the results suggested that HSV latent infection rather than persistent infection was present in transplants. From these results, two explanations seemed possible: latency was maintained in transplanted neurons, or alternatively, latency developed after transplantation, in neurons not previously latently infected. The latter was considered putative secondary latency and was investigated in three ways. First, evidence of reactivation which might serve as a source for secondary latency was evaluated. Reactivation of HSV in transplants was evident from HSV antigen expression (52% of transplants) and the presence of cell-free virus (38% of transplants) 3 to 5 days after transplantation. Second, putative secondary latency was investigated in recipients immunized with HSV prior to receiving latently infected ganglia. Reactivation was not detected 3 to 5 days after transplantation in immunized recipients, and LAT expression was rare in these recipients after 3 to 4 weeks. Lastly, the possibility of secondary latency was investigated by comparing results obtained with standard HSV and with reactivation-defective thymidine kinase-negative (TK-) HSV. Defective reactivation of TK- HSV was demonstrated by immunohistochemistry and by the inability to isolate infectious virus. Donor dorsal root ganglia latently infected with TK+ HSV showed many LAT-positive neurons 2 or more weeks after transplantation (average, 26 per transplant). However, LAT expression was undetectable or minimal > 2 weeks after transplantation in donor ganglia latently infected with TK- HSV (average, 0.2 per transplant). Impaired reactivation of TK- HSV-infected donor ganglia after transplantation, therefore, was correlated with subsequent limited LAT expression. From these results, the occurrence of secondary latency was concluded for ganglia latently infected with TK+ HSV and transplanted beneath the kidney capsule. In vivo reactivation in this transplant model may provide a more useful means to investigate HSV reactivation than in usual in vitro explant models and may complement other in vivo reactivation models. The occurrence of secondary latency was unique. The inhibition of secondary latency by the immune system may provide an avenue to evaluate immunological control of HSV latency.