We evaluated the clinical usefulness of ESR, CRP, and disease duration in ankylosing spondylitis (AS) disease severity. There were 156 Chinese AS patients included in Taiwan. Patients completed the questionnaires, containing demographic data, disease activity (BASDAI), functional status (BASFI), and patient's global assessment (BASG). Meanwhile, patient's physical mobility (BASMI) and acute-phase reactants, including ESR and CRP levels were measured. Receiver operating characteristic (ROC) plot analysis was used to evaluate the performance of ESR, CRP, and disease duration in the AS patients. ESR mildly correlated with BASFI (r = 0.176, p = 0.028) and disease duration (r = 0.214, p = 0.008), and moderately correlated with BASMI (r = 0.427, p < 0.001). CRP moderately correlated with BASMI (r = 0.410, p < 0.001). By using ROC plot analysis, ESR, CRP, and disease duration showed the best and significant "area under the curve (AUC)", in distinguishing the AS patients with poor physical mobility (BASMI ≥ 3.6, the Median) (AUC = 0.748, 0.751 and 0.738, respectively, all p < 0.001), as compared to BASDAI, BASFI, and BASG. ESR × disease duration (AUC = 0.801, p < 0.001) and CRP × disease duration (AUC = 0.821, p < 0.001) showed higher AUC values than ESR or CRP alone in indicating poor physical mobility. For detecting poor physical mobility (BASMI ≥ 3.6) in the AS patients: ESR × disease duration (≥60.0 mm/h × year): sensitivity = 72.7 % and specificity = 72.8 %; CRP × disease duration (≥8.3 mg/dl × year): sensitivity = 72.7 % and specificity = 74.6 %. ESR, CRP, and disease duration are particularly related to AS patient's poor physical mobility. Combining the usefulness of acute-phase reactants and disease duration, the values of ESR × disease duration and CRP × disease duration demonstrate better association with poor physical mobility in AS patients.