Background: The glomerular filtration rate (GFR) is essential for calculating the dose and the monitoring of carboplatin. Although GFR measurement (mGFR) by external markers is ideal, in most cases these are not employed; the most used method is GFR estimation (eGFR) by formulae, hence the need to identify the formula with the best performance. Methods: Patients admitted between 2011 and 2017 and diagnosed with ovarian, endometrial, lung, esophageal, or testicular cancer were assessed retrospectively. The accuracy of the carboplatin dose calculated by creatinine concordance and by the Cockroft-Gault (CG), CKD-EPI, MDRD, Wright, and Jelliffe formulae was assessed using the intraclass correlation coefficient. Results: Fifty-six medical histories were analyzed. The best accuracy was observed between the Wright formula (i.e., 0.71) and the dose calculated based on the 24-h creatinine clearance. Stratification by CKD was made in depurations < 60 mL/min, where the Jelliffe value was excellent (i.e., 0.75). In depurations ≥60 mL/min, CKD-EPI was the best formula, with an accuracy of 0.65. CG was the formula with the worst performance in calculating the dose and glomerular filtration, losing its usefulness with very low filtrations. Conclusions: GFR formulae and calculation of the carboplatin dose have good accuracy with the GFR obtained based on the 24-h creatinine clearance, with the Wright formula being the one with best performance and CG the one with worst performance.