Abstract: OBJECTIVE To evaluate the efficiency of tongue swab PCR testing in diagnosis of the patients with suspected pulmonary tuberculosis and observe the impact of sputum bacterial load and sampling time on the testing efficiency. METHODS Totally 185 patients with suspected pulmonary tuberculosis who were hospitalized in the tuberculosis department of The Fourth People's Hospital of Nanning from Sep. 2025 to Feb. 2026 were prospectively enrolled in the study. Set the sputum Xpert as the reference standard, the sensitivity, specificity and accurate rate of the tongue swab testing were calculated. 95%CI was calculated by Clopper-Pearson method, the consistency between the two methods was evaluated through Kappa consistency check. The subgroup analysis was performed based on the semiquantitative result of sputum Xpert (high, medium, low, and extremely low), and the impact of the bacterial load of sputum on the diagnostic efficiency was observed. The bacterial load of sputum specimens was stratified according to the sampling time (morning vs. daytime), and the impact of the sampling time on the detection performance was evaluated. RESULTS The overall sensitivity of the tongue swab testing was 69.23%(45/65,95%CI: 57.28% to 79.14%), with the specificity 94.17%(113/120,95%CI: 88.42% to 97.16%),Kappa=0.66(95%CI: 0.55 to 0.78). The sensitivity declined remarkably with the decrease of bacterial load, with the sensitivity 93.33%(14/15) in the high bacterial load group, 83.33%(15/18)in the medium bacterial load group, 61.90%(13/21) in the low bacterial load group, 11.11%(1/9) in the extremely low bacterial load group. The sensitivity reached up to 100.00% (9/9) in the morning sampling subgroup (24 cases), with Kappa=0.91,however, the 95%CI was broad (70.09% to 100.00%). CONCLUSIONS The tongue swab PCR testing exhibits high specificity and moderate sensitivity for the patients with suspected pulmonary tuberculosis. The testing efficiency is remarkably affected by the bacterial load of sputum. It can be used for the supported screening of the patients with high bacterial load and the patients who are difficult to obtain the sputum. The pulmonary tuberculosis can not be excluded completely by the negative result when it is low bacterial load, it is necessary to confirm with other methods. The morning sampling may raise the diagnostic efficiency, but it requires a large size of samples for validation.