Electron paramagnetic resonance (EPR) oximetry measurements have provided valuable information about oxygen's distribution, concentration, and diffusion in membranes by using spin-label probes placed at various depths within the bilayer. Nearly all of the probes are confined to relatively narrow regions and their positions are considered well-established. Through molecular dynamics simulations, however, we have found the position of the headgroup probe, tempocholine (T-PC), to be ambiguous. Measuring oxygen in this region is essential to the accurate estimation of membrane oxygen permeability because the resistance to permeation is highest in the headgroup. We have used molecular dynamics simulations to model the T-PC probe and its behavior within 1-palmitoyl,2-oleoylphosphatidylcholine (POPC) bilayers, with and without cholesterol. Simulations with two distinct force-field parameter sets indicate that the T-PC probe resides primarily in the hydrophobic region of the POPC bilayer, just below the polar headgroup. When cholesterol is included in a 1:1 ratio with POPC, we find that the T-PC probe samples the headgroup region and the subheadgroup region, but it does not penetrate into the hydrophobic core because of the rigidity conferred by cholesterol. Taking this behavior into account enables reinterpretation of published EPR data. Our simulations largely agree with those data, while suggesting that the permeability of phospholipid bilayers is substantially lower than estimated in the EPR studies, due to missampling of the headgroup region O2 populations.