The reliable identification of field halo stars born in globular clusters (GCs) is important to understand the actual contribution of GC stars to the Milky Way halo. Depending on the GC formation mechanism, the contribution may be  minimal,  a mere few per cent, or very important, with more than 50\% (see e.g.  the review in Gratton et al. 2012). I will discuss the possibility of finding stars  lost by GCs by chemical tagging, i.e. using the peculiar chemistry found in GCs  focussing on what has been done with "private" data and mini-surveys (e.g., the FLAMES survey of Galactic GCs done by our group, see Carretta et al. 2009a,b; and Bragaglia et al. 2015 for the last results) and what can be obtained from large scale  observational campaigns, such  as the public high-resolution survey  Gaia-ESO or future surveys.