The aim of this prospective study was to evaluate anatomical-functional image fusion using the new technology of combined transmission and emission tomography (SPET/CT) in patients with neuroendocrine tumours (NET). Fifty-four patients with known or suspected NET prospectively underwent both tumour scintigraphy with (111)In-octreotide (n=43) or (123)I-MIBG (n=11) and contrast-enhanced high-end spiral CT. Scintigraphy was performed using a gamma camera (Millennium VG & Hawkeye, GE) with an integrated X-ray tube for combined transmission and emission tomography. SPET and high-end CT were interpreted blinded with regard to localisation and classification of lesions. Analysis of fused images (SPET/CT) was done on a lesion-by-lesion basis, followed by re-evaluation of SPET and high-end CT by consensus. The standard of reference for confirming the presence or absence of malignancy was either histopathology or clinical and imaging follow-up data. A total of 120 lesions were identified by CT and/or scintigraphy. This group included four patients with negative SPET but eight liver lesions on CT that were proven to be metastases. We excluded from the analysis two patients with no evidence of tumour on either modality, two lesions that lacked comparison with the standard of reference and two patients, each with two lesions, who were lost to follow-up. In 56 of the 114 evaluated lesions (49%), the results of SPET and CT were concordant; all lesions were interpreted as malignant. In 58 of 114 lesions (51%), consensus reading of fused images changed the image interpretation of 39 CT scans and 19 SPET studies: 31 lesions previously interpreted as equivocal (n=10) or benign (n=21) were re-classified as malignant and 27 lesions previously interpreted as equivocal (n=19) or malignant (n=8) were re-evaluated as benign. The highest accuracy (99%) in classifying NET lesions was achieved by combined analysis of SPET/CT ("hawkeye") and high-end CT. The specificity of SPET/CT was significantly higher than that of CT alone (P=0.0026) and slightly higher than that of SPET alone, but the accuracy of SPET/CT was inferior to that of side-by-side analysis of SPET and high-end CT (P=0.013) or visual correlation of SPET/CT and high-end CT (P<0.0001). Therapy was changed in 14 of 50 patients (28%) owing to the results of image fusion: in five patients tumour could be excluded, three patients were spared unnecessary surgery because of additional lesions indicating systemic tumour spread, in four patients the surgical approach was changed owing to precise tumour localisation and minimising of the surgical field, and in two patients medical and radiopeptide therapy was changed. Anatomical-functional image fusion allows for improved localisation and characterisation of NET with resultant alteration of the treatment approach in a substantial proportion of patients.