We thank Dr Eisenhut for his interest in our prospective 12-year study of herpes simplex encephalitis in children.1 Because of the standardized comprehensive prospective microbiologic investigations performed on all children who were admitted to our institution with acute encephalitis, evidence of infection with >1 potential pathogen is often observed. This approach is very different from retrospective reviews of cases in which the final diagnosis is the point of entry into the study. In a recent review of our data, approximately one fourth of the cases showed evidence of infection (of variable strength) with ≥2 pathogens.2 In situations such as this, it is often not possible to determine with certainty which pathogen is the causative agent. Furthermore, in some cases, >1 of the potential pathogens may be playing a role.In regard to the specific cases mentioned by Dr Eisenhut, we disagree with his conclusion that Mycoplasma pneumoniae is the more likely pathogen. We believe that most experts would consider the detection of herpes simplex virus (HSV) in the cerebrospinal fluid (CSF) by polymerase chain reaction (PCR) as strong evidence of causality irrespective of HSV serologic results or microbiologic evidence of infection with other potential pathogens. The failure to detect a rise in complement fixation titer to HSV (cases 7 and 8) between acute and convalescent sera could be a result of the well-known limited sensitivity of this assay.3,4 In regard to case 9, it is possible that the convalescent sample was obtained too early (51 days) for HSV seroconversion to have occurred. In 1 prospective study of HSV seroconversion, the median time between onset of genital tract symptoms and seroconversion was 56 days (interquartile range: 16–121 days).5 In addition, there is a paucity of data on the sensitivity of HSV serology in the setting of HSV encephalitis. We acknowledge that when both HSV and M pneumoniae are detected by PCR in the CSF (cases 7 and 9), it is difficult to assign causality solely to 1 pathogen. To avoid subjective assignment of causality, we set rigid inclusion criteria for herpes simplex encephalitis; these criteria are detailed in the methods section of our article.1Dr Eisenhut refers to “local reactivation” of HSV as a possible explanation for the delayed detection of HSV in the CSF of case patient 9. We do recognize that local reactivation (not causally related to the encephalitis) may be an explanation for some patients with positive HSV CSF PCR results but believe this to be unlikely in this case because of the aforementioned stringent criteria used to define HSV encephalitis. We did exclude several cases in which HSV was detected in the CSF by PCR because of a clear alternative diagnosis or absence of typical CSF, electroencephalographic, or neuroimaging findings, again in accordance with preselected exclusion criteria. As discussed in our article, we support a full course of therapy for possible HSV encephalitis for children in whom HSV is detected in the CSF by PCR irrespective of the suspected cause of the encephalitic process. In addition, it has been shown that the detection of HSV in the CSF by PCR of patients with HSV encephalitis can be delayed.6 Hence, HSV should not be excluded as a possible cause of encephalitis solely on the basis of a single negative CSF PCR assay, particularly if the lumbar puncture was performed during the first 48 to 72 hours of illness.We agree that M pneumoniae is an important cause of acute encephalitis in children and that additional research into the role of this pathogen in neurologic disease is needed. In a previous publication pertaining to M pneumoniae and acute childhood encephalitis, we stratified cases into probable, possible, and indeterminate categories in accordance with the strength of microbiologic evidence implicating M pneumoniae as the cause.7 Accordingly, M pneumoniae was considered to be the probable cause of 6.9% of the acute childhood encephalitis cases. As indicated in that article, cases in which the only evidence of M pneumoniae infection was serologic were often associated with more compelling evidence of infection with other potential pathogens. This observation suggests that many of the positive immunoglobulin M assays for M pneumoniae were probably falsely positive, perhaps as a result of cross-reactivity of M pneumoniae antigens and human brain antigens8–10 or other inherent limitations in the specificity of serologic tests for M pneumoniae.11Determining the etiology and pathogenesis of acute childhood encephalitis remains a major hurdle to the appropriate management of this condition. Ultimately, new innovative diagnostic and treatment strategies will be needed to improve the outcome of this often devastating entity.