Determinations of the possible importance of herpes simplex virus type 1 (HSV) thymidine kinase (TK) expression in the pathogenesis of viral latency depend in part on the use of defined mutants. In a recent study by A. E. Sears, B. Meignier, and B. Roizman (J. Virol. 55, 410-416 (1985], in which they utilized genetically engineered viral recombinants considered to be TK-, the role of HSV TK expression in latency was reported to be minimal. To further investigate this conclusion we intensively studied the TK phenotypes of their M316-2 and M316-10 HSV-1 mutants. TK activity was investigated by phosphorylation of thymidine, by arabinosylthymine (ara-T) inhibition and by virus plaque autoradiography. TK activity of the M316-2 and M316-10 HSV mutants was not detected in 5-min assays (as performed by Sears et al.), but in longer assays substantial activity was apparent. In contrast, in assays of control TK- viruses, activity was minimal or absent at all time points. In ara-T inhibition assays the M316-2 and M316-10 viruses were inhibited more than 10-fold, consistent with viruses of intermediate TK activity. By plaque autoradiography both of these viruses produced plaques which incorporated significant amounts of thymidine. Based on these results we conclude that the M316-2 and M316-10 viruses should likely be considered to express intermediate levels of TK activity. HSV latency results using these mutants may need to be interpreted with this in mind.