Many enveloped viruses enter cells via fusion with the endosomal membrane, raising the question whether entry through the endosomal route confers a fitness advantage over fusion directly at the plasma membrane. We found that influenza A virus fusion at the plasma membrane of A549 cells resulted in a 7.6-fold reduction in productive cell infection, compared to infection through the physiological endosomal route. We hypothesized that this was partially explained by restrictive and permissive membrane factors at the plasma and endosomal membrane, respectively. To test this, we developed a single-viral content mixing assay with plasma membrane vesicles (PMVs), where a fluorescent content marker was loaded into purified PMVs at a quenched concentration, through freeze-thawing. We show that influenza fusion with the plasma membrane is 6-fold less efficient than with liposomes containing the endosome-enriched phospholipid BMP. Incorporating BMP into the PMVs with pre-loaded methyl--cyclodextrin (MCD) did not rescue full fusion, suggesting that the lack of BMP in the plasma membrane is not sufficient to explain decreased fusion with the plasma membrane. Depletion of cholesterol from PMVs enhanced lipid mixing rates and reduced membrane order. Lastly, modest depletion of cholesterol from A549 lung epithelial cells rescued influenza infection through fusion at the plasma membrane by 2.4-fold. We propose that for the endosome-adapted influenza virus, fusion directly at the plasma membrane is restricted by the liquid-ordered-like nature of the plasma membrane, raising the question whether the plasma membrane is broadly less permissive to viral entry than endosomes.