Objectives: (i) To propose regression equations based on anthropometric measures to estimate fat mass (FM) using dual energy X-ray absorptiometry (DXA) as reference method, and (ii) to establish population reference standards for equation-derived FM. Material and methods: A cross-sectional study on 6,713 university students (3,354 males and 3,359 females) from Chile aged 17.0 to 27.0 years. Anthropometric measures (weight, height, waist circumference) were taken in all participants. Whole body DXA was performed in 683 subjects. A total of 478 subjects were selected to develop regression equations, and 205 for their cross-validation. Data from 6,030 participants were used to develop reference standards for FM. Equations were generated using stepwise multiple regression analysis. Percentiles were developed using the LMS method. Results: Equations for men were: (i) FM = −35,997.486 + 232.285 * Weight + 432.216 * CC (R2 = 0.73, SEE = 4.1); (ii) FM = −37,671.303 + 309.539 * Weight + 66,028.109 * ICE (R2 = 0.76, SEE = 3.8), while equations for women were: (iii) FM = −13,216.917 + 461,302 * Weight + 91.898 * CC (R2 = 0.70, SEE = 4.6), and (iv) FM = −14,144.220 + 464.061 * Weight + 16,189.297 * ICE (R2 = 0.70, SEE = 4.6). Percentiles proposed included p10, p50, p85, and p95. Conclusion: The developed equations provide valid and accurate estimation of FM in both sexes. The values obtained using the equations may be analyzed from percentiles that allow for categorizing body fat levels by age and sex.