PEGylated PAMAM dendrimers (PEG-PAMAM) have been extensively studied as versatile vehicles for drug delivery. Nevertheless, little information has been reported regarding the effect of the PEGylation degree on the drug-loading properties of these systems, aimed at maximizing their performance as drug carrier nanocarrriers. In this work, fully atomistic molecular dynamics (MD) simulations were employed to examine the association of methotrexate (MTX) with native and diversely PEGylated PAMAM-G4 dendrimers, using 2 kDa PEG chains with substitution degrees from 25 to 100% and 100:1 drug:dendrimer ratios to mimic experimental conditions of drug excess in saturated solution. MD results regarding complex stoichiometries and preferential binding sites were compared to experimental data retrieved from aqueous solubility profiles and 2D-NOESY experiments showing an outstanding level of agreement. The maximum theoretical drug loading capacity was achieved by the system with 34% PEGylation (42:1) through the simultaneous complexation of MTX within internal PAMAM-G4 branches and external PEG chains. On the other hand, higher PEGylation degrees were found to be detrimental for drug complexation due to PEG chains crowding on the dendrimer surface. These results provide valuable information to design more efficient PAMAM-based drug nanocarriers and explain the positive effect that partial PEGylation exerts on the drug-loading capacity of PAMAM-G4 over native and fully PEGylated systems.