The purpose of this study was to determine the utility of radiologist-performed sonography as the principal modality for parathyroid localization before minimally invasive parathyroidectomy. Both sonography and technetium Tc 99m sestamibi single-photon emission computed tomography (SPECT) are commonly performed during imaging evaluation of patients with primary hyperparathyroidism (HPTH). Sonographic examinations ordered during the study period were performed by 1 author (M.E.T.), and results were immediately reported. Findings of a subsequent Tc 99m sestamibi study were recorded blinded to the sonographic results. The sensitivity and specificity of sonography and Tc 99m sestamibi SPECT were assessed with the use of surgery and pathology reports as a reference standard. The 2007 global Medicare reimbursement rates were used to assess the costs of preoperative localization. Parathyroidectomy was performed in 144 of 172 patients evaluated by both modalities. The sensitivity, specificity, and positive predictive value of sonography for identifying abnormal parathyroid glands were 74%, 96%, and 90%, respectively. Sonography correctly localized a single adenoma or suggested multiglandular disease in 112 of 144 patients (78%). The sensitivity, specificity, and positive predictive value of SPECT were 58%, 96%, and 89%. Technetium 99m sestamibi SPECT correctly predicted an adenoma or multiglandular disease in 88 of 144 patients (61%). Five patients with negative sonographic findings were shown to have uniglandular disease on Tc 99m sestamibi SPECT. Selective use of Tc 99m sestamibi SPECT (ie, when sonographic findings were negative or equivocal) would have decreased the cost of imaging by 53%. Radiologist-performed sonography may potentially be used as a principal imaging modality for patients with HPTH. Selective use of Tc 99m sestamibi in cases with negative or equivocal sonographic findings can decrease the cost of imaging before parathyroid resection considerably.