Purpose: We undertook this study to evaluate the need for diffusion-weighted (DW) magnetic resonance (MR) imaging in detecting hemorrhagic infarction following ovarian torsion.Methods: The study included 14 consecutive patients aged 12 to 74 years (average age, 36 years) with surgical confirmation of ovarian torsion who underwent 1.5-tesla MR imaging. Pathologically, hemorrhagic infarction was found in 7 patients. We retrospectively reviewed signal intensity on T 1 -, T 2 -, and diffusion-weighted images and apparent diffusion coefficients (ADCs) in swollen ovarian stroma.Results: Fallopian tube thickening was seen in all patients. In patients with ovarian cystic lesion, maximum cyst wall thickness was significantly higher in patients with hemorrhagic infarction (mean, 13.5 « 4.1 mm) than those without (mean, 5.0 « 1.0 mm) (P < .05). Signal intensity did not differ significantly on T 1 -weighted, T 2 -weighted, and DW images between patients with and without hemorrhagic infarction. ADCs were significantly lower in patients with hemorrhagic infarction (1.20 « 0.50 [© 10 ¹3 mm 2 /s]) than those without (2.04 « 0.26 [© 10 ¹3 mm 2 /s]) (P < .01). With an ADC threshold of 1.80 [© 10 ¹3 mm 2 /s], sensitivity for hemorrhagic infarction was 0.88 (7 of 8), and specificity was 1.00 (6 of 6).Conclusion: ADC measurements were useful for detecting hemorrhagic infarction in patients with ovarian torsion.