We present a study using the second data release of the GALAH survey of stellar parameters and elemental abundances of 15 pairs of stars identified by Oh et al. They identified these pairs as potentially co-moving pairs using proper motions and parallaxes from Gaia DR1. We find that 11 very wide (>1 pc) pairs of stars do in fact have similar Galactic orbits, while a further four claimed co-moving pairs are not truly co-orbiting. Eight of the 11 co-orbiting pairs have reliable stellar parameters and abundances, and we find that three of those are quite similar in their abundance patterns, while five have significant [Fe/H] differences. For the latter, this indicates that they could be co-orbiting because of the general dynamical coldness of the thin disc, or perhaps resonances induced by the Galaxy, rather than a shared formation site. Stars such as these, wide binaries, debris of past star formation episodes, and coincidental co-orbiters, are crucial for exploring the limits of chemical tagging in the Milky Way.