PURPOSE We aimed to describe ultrasonography (US), computed tomography (CT), and magnetic reso-nance imaging (MRI) findings of focal hypersteatosis (FHS).METHODSWe retrospectively reviewed our database for patients with hypersteatosis. Over a 5-year period (February 2005 to September 2010) a total of 17 321 patients underwent abdominal CT scan and 28 patients were determined to have FHS. All patients had US, CT, and MRI studies. Size, area, and density measurements were performed on CT images. Fat signal percentage (FSP) was measured on T1-weighted in- and out-of-phase gradient-echo images. FHS was defined based on MRI findings, as an area of greater signal drop on out-of phase images compared with the rest of the fatty liver. RESULTSThe period prevelance of focal hypersteatosis was measured as 0.16% over the 5-year period. Cancer was the most common diagnosis (22 of 28 patients, 78.5%), with the breast (32.1%) and colorectal (25%) cancers predominating. FHS was seen in segment 4 (n=26, 92.8%), segment 8 (n=1, 3.6%), and segment 3 (n=1, 3.6%). Shape was nodular in 21 patients (75%), while triangular or amorphous in the remaining 7 patients (25%). FHS was hyperechoic and isoechoic in 5 (17.9%) and 23 (82.1%) patients, respectively. FHS was hypodense on CT of all patients relative to fatty liver. On MRI, the FHS was hyperintense on T1-weighted in-phase images in 17 patients (60.7%). Median liver parenchymal FSP was 21.5% (range, 10%–41.4%) and median FSP of hypersteatotic area was 32.5% (range, 19%–45%).CONCLUSION Focal hypersteatosis is a pseudolesion that can be observed in patients with liver steatosis. It ap-pears hypodense on CT and mostly isoechoic on US relative to fatty liver. It may mimic metastasis in cancer patients with steatosis, due to nodular shape and atypical location. MRI should be used for correct diagnosis in patients with equivocal findings on CT to avoid biopsy.