Water is arguably one of the most important molecules, and consequently is object of research in many scientific fields. Traditionally, the Born-Oppenhemeir approximation and a purely non relativistic hamiltonian are the framework in which theoretical spectroscopists work. These approximations allow a full qualitative understanding of the physics underlying water spectrum. However, it has been shown that for pratical purposes, a more detailed knowledge of water spectrum is required. Also, due to its simplicity, water is a benchmark for theoretical calculations. Császár et al. have already demonstrated the importance of one-electron relativistic corrections. Following their works, two-electrons relativistic corrections have been studied. Those terms are shown to be significant. At the same level of accuracy, corrections to the Born-Oppenheimer approximation are investigated. Several models for including non-adiabatic effects are discussed. However, a systematic study of the water spectrum based on all the corrections so far investigated has still to be performed.