The pioneering observations of Spite & Spite showed lithium abundances in halo dwarfs to be almost uniform, irrespective of metallicity and mass over a range of effective temperatures from $\sim$5600 K up to the main-sequence turnoff. They inferred that the observed abundance was “hardly altered” from that produced in the hot Big Bang. Subsequent efforts have endeavoured to determine how small or large “hardly” could be. Simplistic arguments based on the uniformity of the Spite plateau suggest there should only be a small difference between the Big Bang lithium abundance and the observationally inferred plateau value, whereas more physical lines of reasoning suggest the difference could be more substantial. This review paper discusses observational and theoretical developments.