Concentration-dependent binding to red blood cells is a characteristic of several drugs, complicating the understanding of how pathophysiological factors influence drug behavior. This study utilized user-friendly, physiologically-based pharmacokinetic (PBPK) models to compare concentration-dependent and independent blood-to-plasma drug concentration ratios (B/P), using tacrolimus as a case study. Two models were developed and validated for tacrolimus using clinical data from healthy volunteers; Model 1 accounted for saturable blood binding, and Model 2 used a constant B/P level. The differences between the two models based on the two binding assumptions were also studied across clinically relevant hematocrit (HCT) and dose levels. For intravenous (IV) infusions, varying HCT from 15 to 45% resulted in a predicted difference in the area under the concentration-time curve (AUC) of 6–9% for total drug concentration in blood and 37–39% for unbound drug concentration in plasma. Increasing IV doses increased the predicted differences in blood AUC. For oral dosing to steady state, predicted differences in trough concentrations ranged between 50% and 130%, peak concentrations (78–284%), and AUC (up to 125%) according to HCT, dose, and biological medium, e.g., trough differences ranged from 50% (blood, 5 mg) to 130% (plasma, 10 mg). A hypothetical scenario of tacrolimus dose levels increasing above clinically relevant doses revealed a reducing difference in outcomes between the two binding assumptions. Although PBPK models ignoring concentration-dependent binding may adequately fit observed data, they can necessitate compensatory adjustments in disposition parameters, limiting their ability to predict clinical scenarios beyond the model’s original development settings.