The aim of our prospective study was to evaluate the predictive value of serum procalcitonin (PCT) level in comparison with C-reactive protein level and erythrocyte sedimentation rate for the diagnosis of pulmonary tuberculosis (PTB) on admission and 6 months after the administration of anti-tuberculous chemotherapy (ATCT). Seventy-five adult male patients with active PTB who were mycobacteriologically diagnosed (smear and culture positivity) were examined in this study. As a control group, 75 healthy adult males were enrolled. The measured serum PCT levels were within the normal range both in healthy individuals and in patients 6 months after ATCT. Serum PCT levels had been slightly high on admission in patients with PTB in comparison with controls (P = 0.01) and patients who had ATCT (P = 0.001), and this difference was statistically significant, but the PCT levels of most cases with PTB (58.7%) were below the usual cut-off level (0.5 ng/mL). We conclude from this study that the serum PCT level was not a reliable indicator in the diagnosis of active PTB because of its low sensitivity (41.3%), and in most cases it was not capable of overcoming the cut-off level even if statistically meaningful results were obtained. The PCT test for the presumptive diagnosis of PTB cannot be substituted for microbiological, epidemiological, clinical and radiological data.