The accuracy of technetium-99m human immunoglobulin (HIG) for the detection of chronic osteomyelitis (OM) was compared with white blood cell scintigraphy using 99mTc-labelled monoclonal mouse antibodies (MAB). Seventeen patients suspected of having OM in 20 lesions went through three-phase skeletal scintigraphy, HIG scintigraphy and MAB scintigraphy. The final diagnosis was established by open surgery, histology and bacteriology. Chronic OM was proved in 14/20 lesions. Six of these 14 infections were located in peripheral areas without active bone marrow and 8/14 in central areas with active bone marrow. In peripheral OM, 5/6 with HIG and 6/6 with MAB were true positives. In the central skeleton all 8/8 infections appeared as cold lesions in the MAB study, which were defined as being false negative due to their non-specificity. Using HIG, 5/8 central infections were determined to be truly positive by showing photon-rich lesions. These 5 lesions were located in the hip region and in the pelvis, whereas 3 lesions of the spine were missed. There were no false-positive results in either studies. In conclusion, MAB was superior to HIG in peripheral OM concerning sensitivity, anatomical landmarks and differentiation of soft tissue versus bone infection. In central OM MAB detected all lesions accurately, but no differential diagnosis was possible due to the non-specificity of photon-low areas. In this respect HIG seems to be more specific due to the increased accumulation even in central infection sites.