Technetium-99m sestamibi imaging for parathyroid adenoma localization has been performed using both dual-tracer subtraction and double-phase single-tracer washout techniques. The relative accuracy of these two techniques is uncertain. We have developed a modified imaging technique which combines both approaches and have directly compared them in a series of patients with surgically explored hyperparathyroidism. Initial injection of (99m)Tc-pertechnetate 50 MBq was followed by continuous dynamic imaging of the anterior neck for 30 min. (99m)Tc-sestamibi 1,000 MBq was injected intravenously at the midpoint of the acquisition. Delayed images were performed after 2 h. We blindly reviewed 88 consecutive cases of surgically explored hyperparathyroidism that had undergone preoperative scintigraphic localization with this procedure. Images were reformatted to display subtraction-only, early/delayed sestamibi-only and combined images. Scans were reviewed in random order. Of the 68 cases with solitary parathyroid adenoma, the sestamibi-only images gave correct localization in 49 (72%) while there was a statistically significant improvement in accuracy using the subtraction-only images (58 of 68, 85%, P=0.05) and the combined images (61 of 68, 90%, P=0.0015). Reader confidence was also greater with the subtraction-only and combined images than with the sestamibi-only images. Scan performance with parathyroid hyperplasia was less satisfactory. Although the largest gland was usually correctly identified, hyperplasia was difficult to distinguish from a solitary adenoma. Dual-tracer subtraction parathyroid imaging is superior to double-phase sestamibi-only imaging. The washout data may provide additional information in some cases, however, and an approach that combines both techniques may be optimal.