Non-traumatic femoral head necrosis (FHN) is primarily a disease of the middle-aged adult. Early diagnosis, at a time with lacking or minimal clinical symptoms, is mandatory to consider conservative therapy or joint preserving operations as a therapeutic option. The new German S3 guideline about diagnosis and therapy of FHN is a cooperative effort of five professional medical societies, overall headed by the Deutsche Gesellschaft für Orthopädie und Orthopädische Chirurgie (DGOOC). This review (part I/III) cites and explains the statements of the S3 guideline as agreed on the use of imaging methods for diagnosis of FHN. A diagnostic algorithm is presented. FHN clinically has to be considered in case of equivocal pain of a hip joint with a minimum of 6 weeks duration, when risk factors can be revealed, groin pain at clinical investigation, limping, pain or limitation of movement in case of load, and no obvious differential diagnoses. Is an FHN clinically suspected, primarily radiographs of the pelvis ap and a Lauenstein projection of the hip involved should be carried out. When the radiographs are normal, an MRI of the hips should follow routinely. MRI allows the diagnosis of FNH with high accuracy. Furthermore, MRI reveals the site and the size of the necrotic area involved and evaluates the integrity of the joint surface and subchondral fractures. When ARCO stage II (ARCO: Association Research Circulation Osseous) is diagnosed and MRI does not allow one to determine the joint surface with certainty, a CT of the hip joints should be performed. The S3 guideline explains and recommends the use of the ARCO classification. Although, this classification of 1993 is still largely based on radiographs, the pragmatic use of an "extended" version seems reasonable. Today, classical radiographic criteria like impression of the joint surface and subchondral fractures ("crescent sign") are better to be evaluated by MRI, in cases of subtle findings MRI is even surpassed by CT. The extent of the necrosis in the femoral head as well as the size of the surface area involved is best revealed with MRI. Additionally, in the era of cross sectional imaging a stage "0" seems obsolete. The guideline also addresses practically important considerations about the differential diagnosis of misleading MRI findings. This especially holds true for bone marrow oedema in the femoral head which may be misinterpreted. The differentiating features between FHN, transient bone marrow oedema and destructive arthropathy are discussed.