Transient hyperglycemia is seen commonly during TB treatment, yet its association with unfavorable treatment outcomes is unclear. Does an association exist between glycated hemoglobin (HbA1c) trajectories and TB treatment outcomes? Adults with pulmonary TB were evaluated prospectively for 18months after the second HbA1c measurement. HbA1c trajectories during the initial 3months of treatment were defined as follows: persistent euglycemia, HbA1c< 6.5%at baseline and 3-month follow-up; persistent hyperglycemia, HbA1c≥ 6.5%at baseline and 3-month follow-up; transient hyperglycemia, HbA1c≥ 6.5%at baseline and< 6.5%at 3-month follow-up; incident hyperglycemia, HbA1c< 6.5%at baseline and≥ 6.5%at 3-month follow-up. Multivariable Poisson regression was used to measure the association between HbA1c trajectories and unfavorable treatment outcomes of failure, recurrence, and all-cause mortality. Of the 587 participants, 443 participants (76%) had persistent euglycemia, 118 participants (20%) had persistent hyperglycemia, and 26 participants (4%) had transient hyperglycemia. One participant had incident hyperglycemia and was excluded. Compared with participants with persistent euglycemia, those with transient hyperglycemia showed a twofold higher risk of experiencing an unfavorable treatment outcome (adjusted incidence rate ratio [aIRR], 2.07; 95%CI, 1.04-4.15) after adjusting for confounders including diabetes treatment, and BMI; we did not find a significant association with persistent hyperglycemia (aIRR, 1.64; 95%CI, 0.71-3.79). Diabetes treatment was associated with a significantly lower risk of unfavorable treatment outcomes (aIRR, 0.38; 95%CI, 0.15-0.95). Transient hyperglycemia and lack of diabetes treatment was associated with a higher risk of unfavorable treatment outcomes in adults with pulmonary TB.