Permeapad® is an artificial biomimetic barrier for in vitro permeation experiments, which has an intricate nano- and microstructure consisting of two cellulose hydrate sheets enclosing a layer of phospholipids forming multiple, multilamellar vesicles in contact with the assay medium. Due to this structure, transport across this barrier can be regarded as complex deserving further attention. Until now, only Permeapad® with phosphatidylcholine, the most abundant phospholipid in cell membranes, has been described in literature. However, from biological systems and other artificial barriers, it is known that permeation properties can vary with phospholipid composition. This study presents a combination of experimental and computational techniques to study and explain the transport of molecules across the Permeapad® barrier. For this, we investigated Permeapad® variants with other phospholipid compositions including phosphatidylethanolamine, the second most abundant phospholipid in cell membranes, and phosphatidylglycerol, representing a phospholipid with a negatively charged headgroup by measuring the permeability of three drugs, metoprolol (a weak base), naproxen (a weak acid) and hydrocortisone (a non-ionizable drug). Phospholipid composition only affected the permeability of metoprolol significantly. We used molecular dynamics simulations to understand the underlying mechanisms of the permeability differences extracting several descriptors of membrane properties and predicting permeability. Surprisingly, an almost inverse relationship between experimental and computational permeability was observed. Permeapad®'s highly compartmentalized structure was hypothesized to cause this observation. This study offers a deeper understanding of the functionality of the Permeapad® barrier.