Paired associative stimulation (PAS) is a widely used noninvasive brain stimulation protocol to assess neural plasticity. Its reproducibility, however, has been rarely tested and with mixed results. With two consecutive studies, we aimed to provide further tests and a more systematic assessment of PAS reproducibility. We measured intraclass correlation coefficients (ICCs)-a widely used tool to assess whether groups of measurements resemble each other-in two PAS studies on healthy volunteers. The first study included five PAS sessions recording 10 MEPS every 10 min for an hour post-PAS. The second study included two PAS sessions recording 50 MEPS at 20 and 50 min post-PAS, based on analyses from the first study. In both studies PAS sessions were spaced one week apart. Within sessions ICC was fair to excellent for both studies, yet between sessions ICC was poor for both studies. We suggest that long term meta-plasticity effects (longer than one week) may interfere with between sessions reproducibility.