We present an updated mid-infrared (MIR) versus X-ray correlation for the local active galactic nuclei (AGN) population based on the high angular resolution 12 and 18μm continuum fluxes from the AGN subarcsecond MIR atlas and 2–10 keV and 14–195 keV data collected from the literature. We isolate a sample of 152 objects with reliable AGN nature and multi-epoch X-ray data and minimal MIR contribution from star formation. Although the sample is not homogeneous or complete, we show that our results are unlikely to be affected by significant biases. The MIR–X-ray correlation is nearly linear and within a factor of 2 independent of the AGN type and the wavebands used. The observed scatter is <0.4 dex. A possible flattening of the correlation slope at the highest luminosities probed (∼1045  erg s−1) towards low MIR luminosities for a given X-ray luminosity is indicated but not significant. Unobscured objects have, on average, an MIR–X-ray ratio that is only ≤0.15 dex higher than that of obscured objects. Objects with intermediate X-ray column densities (22 < log N H < 23) actually show the highest MIR–X-ray ratio on average. Radio-loud objects show a higher mean MIR–X-ray ratio at low luminosities while the ratio is lower than average at high luminosities. This may be explained by synchrotron emission from the jet contributing to the MIR at low luminosities and additional X-ray emission at high luminosities. True Seyfert 2 candidates do not show any deviation from the general behaviour suggesting that they possess a dusty obscurer as in other AGN. Double AGN also do not deviate. Finally, we show that the MIR–X-ray correlation can be used to investigate the AGN nature of uncertain objects. Specifically, we give equations that allow us to determine the intrinsic 2–10 keV luminosities and column densities for objects with complex X-ray properties to within 0.34 dex. These techniques are applied to the uncertain objects of the remaining AGN MIR atlas, demonstrating the usefulness of the MIR–X-ray correlation as an empirical tool.