When processed in solid or liquid medium, tuberculosis patient samples yield different proportions of a heterogenous bacterial community over the duration of treatment. We aimed to derive a relationship between methodologies for bacterial load determination and assess the effect of the growth phase of the parent culture and its exposure to stress on the results. Mycobacterium tuberculosis H37Rv was grown with and without antibiotic (isoniazid or rifampicin) and sampled on day 0, 3, 11 and 21 of growth in broth culture. The bacterial load was estimated by colony counts and the BD BACTEC MGIT system. Linear and nonlinear mixed-effects models were used to describe the relationship between time-to-positivity (TTP) and time-to-growth (TTG) versus colony forming units (CFU), and growth units (GU) versus incubation time in MGIT. For samples with the same CFU, antibiotic-treated and stationary phase cells had a shorter TTP than antibiotic-free controls and early-logarithmic phase cells, respectively. Similarly, stationary phase samples reached higher GUs and had shorter TTG than early-log phase ones. This suggests that there is a population of bacterial cells that can be differentially recovered in liquid medium, giving us insight into the physiological states of the original culture, aiding the interpretation of clinical trial outputs.