We discuss the opportunities to use the future ELTs to study in detail the properties and the evolution peculiarities of individual massive stars with a metal content typical of galaxies in the Universe during the first 0.5–1Gyr, corresponding to the earliest known newly formed galaxies at redshifts of $z$ = 6–10. This is possible in principle due to the existence in the local Universe of starbursting galaxies with metallicities of $\sim$1/30 Z$_{\odot}$. The nearest such galaxies are DDO 68 at $\sim$6.5 Mpc with 12+log(O/H) = 7.21 and I Zw 18 at 15 Mpc with 12+log(O/H) = 7.17. For the youngest star clusters (with ages of T $<$ 4–5Myr) in these most metal-poor galaxies, stars with masses up to 40–60 M$_{\odot}$ should be present on both the Main Sequence and the later evolution stages, including the WR stage. They are expected to have apparent magnitudes as bright as $V\,{=}\,21^m$ for DDO 68 and $V\,{=}\,23^m$ for I Zw 18. Good S/N-ratio spectroscopy with telescopes like OWL or JWST, allowing near-milli-arcsecond angular resolution, will provide unique information to check the most up-to-date models of massive star evolution in the very low metallicity regime, and thus, establish a firm basis to model the effects of star formation in ‘primordial’ galaxies, at the epoch of galaxy formation. Such observations will also provide an independent channel to probe the primordial Helium abundance.