The Multi-site All-Sky CAmeRA (MASCARA) aims to find the brightest transiting planet systems by monitoring the full sky at magnitudes $4<V<8.4$, taking data every 6.4 seconds. The northern station has been operational on La Palma since February 2015. These data can also be used for other scientific purposes, such as the study of variable stars. In this paper we aim to assess the value of MASCARA data for studying variable stars by determining to what extent known variable stars can be recovered and characterised, and how well new, unknown variables can be discovered. We used the first 14 months of MASCARA data, consisting of the light curves of 53 401 stars with up to one million flux points per object. All stars were cross-matched with the VSX catalogue to identify known variables. The MASCARA light curves were searched for periodic flux variability using generalised Lomb-Scargle periodograms. If significant variability of a known variable was detected, the found period and amplitude were compared with those listed in the VSX database. If no previous record of variability was found, the data were phase folded to attempt a classification. Of the 1919 known variable stars in the MASCARA sample with periods $0.1<P<10$ days, amplitudes $>2%$, and that have more than 80 hours of data, $93.5%$ are recovered. In addition, the periods of $210$ stars without a previous VSX record were determined, and $282$ candidate variable stars were newly identified. The OConnell effect is seen in seven eclipsing binaries, of which two have no previous record of this effect. MASCARA data are very well suited to study known variable stars. They also serve as a powerful means to find new variables among the brightest stars in the sky. Follow-up is required to ensure that the observed variability does not originate from faint background objects.