Pyrazinamide (PZA) is an important front-line anti-tuberculosis drug that is active only at acid pH. However, acid pH causes significant difficulty for PZA susceptibility testing. A common problem in PZA testing is false resistance caused by large bacterial inocula. This study investigated the relationship of false resistance to numbers of bacilli, pH and other factors that potentially affect susceptibility to PZA. Large inocula (10(7-8) bacilli/ml) of M. tuberculosis H37Ra caused significant increase in medium pH from 5.5 towards neutrality, and thus produced false resistance results. The increase in medium pH was determined to be a function of live bacilli; heat-killed bacilli had little or no effect. Susceptibility to PZA and its active derivative pyrazinoic acid (POA) was comparable on 7H11 agar medium, but POA was less active than PZA in liquid medium containing bovine serum albumin (BSA), suggesting that susceptibility to PZA or POA was reduced in the presence of BSA, because of its neutralising effect on medium pH and significant POA binding. A 3-month-old H37Ra culture was shown to be more susceptible to PZA exposure than a 4-day log-phase culture, suggesting that PZA is more active for non-growing bacilli. Finally, reserpine, an inhibitor of POA efflux pump, increased susceptibility to PZA even near neutral pH 6.8, with an MIC of 400 mg/L compared with 1,000 mg/L without reserpine. These findings should have implications for understanding the mode of action of PZA and for PZA susceptibility testing.