Reliability of mean sleep latency testing (MSLT) over consecutive days in patients with hypersomnia is unknown. We reviewed MSLTs of patients with hypersomnia without cataplexy who underwent our two consecutive MSLT protocol (N=29). Average MSLs were 10.9 and 10.9 minutes for days 1 and 2, respectively. Agreement for pathological hypersomnia (defined as MSL≤8 minutes) between MSLT days showed k=0.85 for all (N=29) and k=0.76 for those without sleep apnea (N=20). In patients with subjective complaints of hypersomnia, a single MSLT is sufficient (vs. addition of 2nd day MSLT) in the setting of carefully implemented protocol controlling for potential confounding variables.