Surrogate markers for monitoring immuno-virological discordant responders, in addition to plasma viral load and CD4 cells, are still lacking. We assessed the diagnostic utility of CD38 expression on CD8 T cell assay, alone or in association with lymphocyte proliferation to mycotic antigens, in evaluating antiretroviral response. 28 vertically HIV-infected youths, 21 HAART- and seven 2 nucleotide reverse transcriptase inhibitors-treated, were enrolled in a retrospective study. Responders (57.1%) and non-responders (42.9%) to stable antiretroviral therapy for a minimum of 6 months, on the basis of viral load and CD4 T cells, comprehensively evaluated by CD38 expression on CD8 T lymphocytes [measured as CD38 antibody bound per CD8 T cell (CD38 ABC) and %CD38+ of total CD8 T cells (%CD38/CD8)] and lymphocyte proliferation to P. jiroveci, C. albicans, C. neoformans, A. fumigatus at a single time point after treatment, were selected. CD38 expression > or =2401 CD38 ABC and > or =85% CD38/CD8 cut-off points, accurately discriminates responders versus non-responders, both measures resulting in 75.0% (CI 42.8-94.5) sensitivity (identification of non-responder) and 93.8% (CI 69.8-99.8) specificity (identification of responder), when considered as single assays. The association '> or =2401 CD38 ABC or > or =85% CD38/CD8' improved sensitivity to 83.3% (CI 51.6-97.9), while the association '<2401 CD38ABC (or <85% CD38/CD8) and lymphoproliferative response positive to > or =2 tested organisms' improved specificity to 100% (CI 79.4-100). In conclusions, CD38 expression and mycotic antigen-specific T-cell proliferation may be used as additional parameters to existing criteria to evaluate antiretroviral response in immuno-virological discordant patients.