Introduction: Plasma is the standard biological fluid used in pharmacokinetic (PK) studies and therapeutic drug monitoring (TDM) of pyrazinamide, a key antituberculosis (TB) drug. This study described the PK of pyrazinamide in saliva and investigated whether saliva could serve as an alternative matrix for pyrazinamide PK evaluations. Methods: Fifteen adult Tanzanian TB patients in the intensive treatment phase participated in a descriptive PK study. Time-matched saliva (stimulated using a Salivette^® with citric acid) and plasma samples were collected at multiple intervals up to 24 h after drug intake. Pyrazinamide concentrations were measured using validated HPLC methods, exposure measures were assessed, and predictive performance for salivary concentrations was determined. Results: Salivary exposure to pyrazinamide (AUC_0–24h: 230 h·mg/L; C_max: 28.6 mg/L) was lower than plasma exposure (AUC_0–24h: 377 h·mg/L; C_max: 36.4 mg/L, p < 0.001), but T_max was similar (median 2.0 h, p = 0.893). A saliva/plasma ratio of 0.59 was assessed, and a reciprocal conversion factor of 1.68 allowed for reasonably accurate (bias 5.8%) but imprecise (imprecision 24.3%) plasma concentration predictions from saliva. Use of a conversion factor of 1.49, based on more stable saliva/plasma concentration ratios for samples between 2 and 6 h post-dose, resulted in a bias of 0.74% and imprecision of 17.7% for predicting plasma concentrations from salivary concentrations in the 2–6 h interval. Conclusions: The exposure to pyrazinamide in saliva is relatively high. Salivary measurement of pyrazinamide can be used as a semi-quantitative predictor of pyrazinamide plasma concentrations.