Star-forming systems are now used to study the universe at z>5, but these galaxies represent <3% of the total baryonic mass. In a pilot programme, we detected CO(2-1) emission at z=5.1245+/-0.0001 (with no optical counterpart to R>28 and I>27) in a field hosting an overdensity of UV-luminous star-forming Lyman break galaxies (LBGs) at the same redshift. The emission line has a peak flux density of 0.94 mJy and FWHM of 110 km/s. Assuming standard conversion factors and that the line width represents a virialised system, this implies M(H_2)=2x10^{10} solar masses, more the total UV-bright stellar mass in the structure. This detection implies that CO lines may probe the bulk of the baryonic matter in z>=5 structures. We must now determine whether our initial detection was typical and intend to study the cool gas in a second field which contains the richest overdensity in our LBG survey. To this end, we will probe an additional three pointings in the CO(2-1) line, and also begin to constrain the temperature from CO(1-0) emission at the same locations.