The abundance of water vapour in molecular clouds is one of the key problems in the study of the interstellar medium (ISM). Although chemical models predict high water abundance, ground-based observations of non-masering rotational transitions of water are nearly impossible due to the terrestrial atmospheric opacity, even from airborne platforms. Cernicharo et al (1994) reported widespread water vapour emission in Orion at 183 GHz and derived water abundances in excess of 10(-5), confirming that H2O is the dominant coolant in warm molecular clouds. Recent observations using the LWS grating have lead to spectacular results concerning the water content in the interstellar medium. The discovery of absorption in the 179.5 micron fundamental transition of water vapour has been reported towards galactic compact HII regions (Cox et al. 1996) and towards the strong continuum of Sgr B2 (Cernicharo et al. 1996). Similar results have also been found for normal galactic lines of sights by Caux et al. (1996). The absorption likely arises from the gas surrounding the central source and indicate the presence of water in the associated molecular envelope and/or along the lines of sight in the normal interstellar medium. Conservatives lower limits of the water abundance of 10(-6) are derived from these very opaque transitions which are readily detected with the grating spectrometer. We propose to measure the compact HII regions for which we have evidence of water absorption by using the LWS Fabry-Perot and the isotope H18O in order to investigate the abundance of water more precisely. ISO is the only instrument in the next decade which will allow to observe at high spatial resolution and sensitivity the thermal emission of water vapor in the ISM. These follow up observations together with similar proposals on the study of water emission in molecular clouds could be one of the major contributions from ISO and without doubt provide a significant progress to our undestanding of the role of water vapour in the chemistry and physics of the ISM.