We performed a prospective study to evaluate the imaging potential of thallium-201 as compared with other imaging modalities in differentiating residual/recurrent tumors from post-therapy changes in patients with musculoskeletal sarcomas. Tl-201 scans, magnetic resonance imaging (17), X-ray computed tomography (6) or contrast angiography (6) studies in 29 patients previously treated for musculoskeletal sarcomas were correlated with either histopathologic findings (26 patients) or 2-year clinical follow-up (three patients). All imaging studies were acquired within 2 weeks. Ratios of Tl-201 tumor uptake to the contralateral (28 patients) or adjacent region of interest were calculated. When qualitative interpretation was in doubt, only those cases with a ratio of 1.5 or more were considered suggestive of re current or residual viable tumor tissue. Residual or recurrent tumor tissue was verified in 21 patients by biopsy, All had true-positive Tl-201 scans while the other imaging modalities were true-positive in 20 and equivocal in one. In eight patients, there was no evidence of viable tumor tissue as proven by biopsy in five and long-term clinical follow-up in three. Tl-201 scan was false-positive (ratio 1.5) in one patient and true-negative in seven while the other imaging modalities had four false-positives. The average Tl-201 ratios were 3.8+/-1.1 in the true-positive cases and 1.3+/-0.3 in the true-negative cases, The percentage sensitivities, specificities, and accuracy for Tl-201 were 100%, 87.5%, and 96.5% versus 95%, 50%, and 82.7% respectively for other imaging modalities, These results indicate that Tl-201 scintigraphy is more accurate than other imaging modalities in differentiating residual/recurrent musculoskeletal sarcomas from post-therapy changes.