In pediatric PBPK models, age-related changes in the body are known to occur. Given the sparsity of and the variability associated with relevant physiological parameters, different PBPK software providers may vary in their system’s data. In this work, three commercially available PBPK software packages (PK-Sim®, Simcyp®, and Gastroplus®) were investigated regarding their differences in system-related information, possibly affecting clearance prediction. Three retrograde PBPK clearance models were set up to enable prediction of pediatric tramadol clearance. These models were qualified in terms of total, CYP2D6, and renal clearance in adults. Tramadol pediatric clearance predictions from PBPK were compared with a pooled popPK model covering clearance ranging from neonates to adults. Fold prediction errors were used to evaluate the results. Marked differences in liver clearance prediction between PBPK models were observed. In general, the prediction bias of total clearance was greatest at the youngest population and decreased with age. Regarding CYP2D6 and renal clearance, important differences exist between PBPK software tools. Interestingly, the PBPK model with the shortest CYP2D6 maturation half-life (PK-Sim) agreed best with the in vivo CYP2D6 maturation model. Marked differences in physiological data explain the observed differences in hepatic clearance prediction in early life between the various PBPK software providers tested. Consensus on the most suited pediatric data to use should harmonize and optimize pediatric clearance predictions. Moreover, the combination of bottom-up and top-down approaches, using a convenient probe substrate, has the potential to update system-related parameters in order to better represent pediatric physiology.