Abstract Aims To perform evaluation of widely embraced bone scintigraphy-based non-biopsy diagnostic criteria (NBDC) for ATTR amyloid cardiomyopathy (ATTR-CM) in clinical practice, and to refine serum free light chain (sFLC) ratio cut-offs that reliably exclude monoclonal gammopathy (MG) in chronic kidney disease. Methods and results A multi-national retrospective study of 3354 patients with suspected or histologically proven cardiac amyloidosis (CA) referred to specialist centres from 2015 to 2021; evaluations included radionuclide bone scintigraphy, serum and urine immunofixation, sFLC assay, eGFR measurement and echocardiography. Seventy-nine percent (1636/2080) of patients with Perugini grade 2 or 3 radionuclide scans fulfilled NBDC for ATTR-CM through absence of a serum or urine monoclonal protein on immunofixation together with a sFLC ratio falling within revised cut-offs incorporating eGFR; 403 of these patients had amyloid on biopsy, all of which were ATTR type, and their survival was comparable to non-biopsied ATTR-CM patients (p = 0.10). Grade 0 radionuclide scans were present in 1091 patients, of whom 284 (26%) had CA, confirmed as AL type (AL-CA) in 276 (97%) and as ATTR-CM in only one case with an extremely rare TTR variant. Among 183 patients with grade 1 radionuclide scans, 122 had MG of whom 106 (87%) had AL-CA; 60/61 (98%) without MG had ATTR-CM. Conclusion The NBDC for ATTR-CM are highly specific [97% (95% CI 0.91-0.99)] in clinical setting, and diagnostic performance was further refined here using new cut-offs for sFLC ratio in patients with CKD. A grade 0 radionuclide scan all but excludes ATTR-CM but occurs in most patients with AL-CA. Grade 1 scans in patients with CA and no MG are strongly suggestive of early ATTR-type, but require urgent histologic corroboration.