The study investigated whether ultrasound-detected fat fraction (UDFF) could be used to predict the severity of PCOS in obese women. The research included 103 women with PCOS and obesity and 108 women with only obesity without PCOS. Women with PCOS and obesity had significantly higher values of UDFF, number of follicles, LH hormone levels and indicators of ovarian perfusion compared to the control group. Logistic regression analysis revealed that UDFF together with other factors (SWV, VFI, RI and LDL) are independent risk factors for PCOS with obesity, with a prediction accuracy of 91.3%. UDFF showed correlation with glucose and lipid metabolism, sex hormone levels, ovarian morphology and hemodynamic parameters. The results suggest that UDFF may serve to predict the severity of PCOS in obese women to some extent.