Background: The triglyceride–glucose index (TyG) and a combination of body mass index (BMI) and waist circumference (WC) have been proposed as predictive scores for liver steatosis (LS). The aim of this study was to determine the diagnostic accuracy of these indices compared with controlled attenuation parameters (CAPs) and other predictive scores of LS. Methods: A retrospective analysis of patients who attended a check-up unit in 2021 was performed. LS was determined by CAP. Anthropometric and biochemical parameters for calculating TyG, TyG-BMI, TyG-WC, fatty liver index, and hepatic steatosis index were obtained. ROC curve was used to establish the best cut-off point of each TyG index for LS detection. The accuracy was determined for all patients, as well as for overweight and diabetic patients. Results: Medical records of 855 patients with a median age of 48 [IQR, 44–54] years and a BMI of 25.7 [IQR 23.4–28.1] kg/m2 were included. According to CAP, LS prevalence was 31.8% (n = 272). TyG-BMI and TyG-WC show better AUCs compared with CAP (0.82, 0.81), FLI (0.96, both), and HSI (0.93, 0.85). For diabetic patients, TyG-WC shows an AUC of 0.70. Meanwhile, TyG-BMI shows better accuracy (0.75) compared with CAP. Conclusions: TyG-BMI and TyG-WC showed a superior predictive accuracy for detecting LS compared with the TyG index.