In vitro stimulation of whole blood or isolated peripheral blood cells with specific antigens is used for several purposes. We sought to identify a reliable, reproducible, fast and feasible in vitro method to assess human cellular immune responses to Mycobacterium tuberculosis. In contrast to peripheral blood mononuclear cell (PBMC) culture, a whole blood assay (WBA) provides a more physiological environment, which may provide a broader assessment of serum biomarker, biosignature profiles. Twenty-three asymptomatic individuals with M. tuberculosis infection were recruited. Total cells from the WBA (diluted 1:3 in completed RPMI) and PBMC (2 × 105 cells/ml) plus M. tuberculosis Ag85A, Ag85B, ESAT-6 and Mycobacterium bovis 65 kDa were characterized by flow cytometry, then added in 96-well plates and on day 5 plasma and supernatants were harvested for detection of 17 cytokines by a Luminex array system. There was agreement between PBMC and WBA for IL-2, IL-5, IL-6, IL-7, IL-13, IFN-γ, TNF-α, MCP-1 and MIP-1β. There was evidence toward higher IL-10 (p ≤ 0.049) and G-CSF (p ≤ 0.012) plasma production, and higher IL-1β (p ≤ 0.048), IL-4 (p ≤ 0.044), IL-12p70 (p ≤ 0.006), IL-17 (p ≤ 0.002) and GM-CSF (p ≤ 0.049) production for PBMC vs. WBA. Both methods provided virtually no reaction to the internal, negative control. Due to technical issues linked to data out of range, IL-8 data were not considered. These results suggest that, depending on the method employed, PBMC and/or WBA techniques provide fine conditions for the model proposed and thus whole blood cultures are well-suited low-cost proxy-measures during search for serum biomarkers.