An accurate estimated glomerular filtration rate (eGFR) is essential in drug dosing. This study demonstrates the limitations of indexed (ml/min/1.73 m2) and de-indexed (ml/min) eGFR based drug dosing in patients with obesity or underweight. This systematic study aimed to determine the most appropriate approach to estimate the GFR for standardized eGFR based drug dosing in these patients. (Raw) data of 12 studies were selected to investigate the accuracy and bias of both the indexed and de-indexed estimations of the Modification of Diet in Renal Disease (MDRD) study equation and the Chronic Kidney Disease Epidemiology Collaboration equation (CKD-EPI), and of the Cockcroft–Gault (CG) in patients with obesity or underweight. Accuracy was calculated as the proportion of eGFR values within 30% of the measured GFR (P30) using an inert tracer (e.g., iohexol, inulin, 51Cr-EDTA, or iothalamate clearance). An accuracy of at least 80% was considered acceptable. GFR values estimated with the CG, MDRD, and CKD-EPI differ significantly within a patient with obesity or underweight regardless of whether it is indexed or de-indexed. All studies, with two exceptions, show that all three equations are inaccurate for patients with underweight or class II obesity (P30: 55%–94%). De-indexing eGFR improves not or modestly the accuracy, and mostly remains below the 80% (P30: 62%–100%). CG was highly inaccurate in obese and underweight patients (P30: 7%–82%). Although these results show that CG is obsolete, the accuracy of MDRD and CKD-EPI is low in patients with obesity or underweight and de-indexing is not the solution. Better education and more accurate methods for appropriate drug dosing (e.g., measured GFR with inert tracer, therapeutic drug monitoring, or 24-h creatinine clearance) are recommended.