In this issue of the European Journal of Nuclear Medicine and Molecular Imaging, Flotats et al. [1] report a striking improvement in image quality of myocardial perfusion studies with the use of 82 Rb PET/CT as compared to conventional 99m Tc-labelled SPECT/CT. When using PET/CT in concert with the positron-emitting myocardial perfusion tracer 82 Rb, the improvement in image quality also manifested in a higher interpretative confidence and interreader agreement of 82 Rb perfusion images than observed for 99m Tc-labelled SPECT/CT. In particular, the current study is unique in that 82 Rb PET/CT and 99m Tc-labelled SPECT/ CT were performed in the same patient with known or suspected coronary artery disease (CAD). Previous investigations commonly compared the sensitivity and specificity of cardiac PET versus SPECT perfusion imaging in the detection of flow-limiting epicardial lesions [2]. For example, Go et al. [3] showed sensitivity, specificity and accuracy of 95, 82 and 92 % for 82 Rb PET and 79, 76 and 78 % for 201 Tl SPECT for haemodynamically obstructive CAD lesions. Further, Stewart et al. [4] reported that overall sensitivity, specificity and accuracy of 82 Rb PET for detection of coronary artery lesions ≥50 % diameter stenosis were 84, 88 and 85 %, respectively. In comparison to this, the performance of 201 Tl SPECT revealed a sensitivity of 84 % but a low specificity of 53 % and a diagnostic accuracy of 79 %. The commonly higher sensitivity in the identification of flow-limiting epicardial lesions with 82 Rb PET as compared with SPECT imaging, either with 201 Tl or 99m Tc-labelled perfusion tracers, can be related to the higher spatial and contrast resolution of 82 Rb PET, while the photon attenuation-free images of PET imaging may account for the relative increase in specificity [2].The new investigation conducted by Flotats et al.[1] now adds further important information by addressing image quality of myocardial perfusion studies, and the resulting reader confidence, and interreader agreement when applying 82 Rb PET/CT in direct comparison to 99m Tc-labelled SPECT/CT. As it was observed, image quality was estimated to be excellent or good in 63 % of PET, 33 % of attenuationcorrected (AC) SPECT and 22 % of non-AC SPECT images. Further, artefact-free images were more frequently observed with PET than with SPECT studies (PET 81 %, AC SPECT 22 % and non-AC SPECT 15 %). In this direction, significant liver or bowel uptake (or both) as well as attenuation artefacts can be assumed to have substantially affected the interpretation of the SPECT/CT images rendering the interpretative decision-making process more difficult or observer dependent for SPECT/CT than for PET/CT images. Not surprisingly, the reader certitude to grade stress-rest perfusion imaging into a normal or abnormal finding was highest for PET (85 %), followed by AC SPECT (30 %) and non-AC SPECT (7 %). This is also reflected by the interreader agreement which was superior for PET/CT than for those of SPECT/CT perfusion studies.Several methodologi...