We prospectively studied the diagnostic potential of Tl-201 and Tc-99m-sestamibi (MIBI) SPECT for evaluating the extent of primary disease and differentiating residual/recurrent disease from posttherapy changes in patients with nasopharyngeal carcinoma (NPC). Methods: Fifty patients (20 initial presentation, 30 post-therapy evaluation) underwent (TI)-T-201 and MIBI imaging. The findings were correlated with CT/MRI results. Tumor-to-background ratios were obtained. Biopsy confirmation (14 patients) and/or 6-12 mo clinical follow-up data (16 patients) were available in the post-therapy group. Results: All primary disease sites were accurately detected by both imaging studies in the pretherapy group. However, MIBI-SPECT was superior to (TI)-T-201 SPECT (p = 0.0057) in detecting regional metastases (sensitivities of 95% versus 68%). In the posttherapy group, MIBI and (TI)-T-201 imaging were true-positive in 14 of 16 patients with proven residual/recurrent. In 17 patients who had no evidence of residual/recurrent tumor, CT/MRI was false-positive in 13 when MIBI and (TI)-T-201 imaging were true-negative in 10 and false positive in 3. MIBI, (TI)-T-201 and CT/MRI had sensitivities of 87.5%, 87.5%, 100%, specificities of 82.4%, 76.5%, 23.5% and accuracies of 85%, 82%, 61%, respectively. Tumor-to-background ratios were less than or equal to 1.5 in all false-positive cases except one. Conclusion: MIBI-SPECT proves more accurate than (TI)-T-201 SPECT in detecting regional metastases at initial presentation. MIBI and (TI)-T-201 imaging have higher specificity and accuracy than CT/MRI and MIBI-SPECT is slightly more specific than (TI)-T-201 SPECT in differentiating residual/recurrent disease from post-therapy changes in patients with NPC.