Switch on the weather forecast any evening and one is left in no doubt as to the importance of water in the Earth's atmosphere: will the water be gaseous (humidity), liquid (clouds, rain), solid (snow, ice or hail) or taken up by aerosol (haze) particles? The long-term role of water in our climate, and of course any change in our climate, is no less profound. The water vapour is both the dominant absorber of incoming sunlight and the major greenhouse gas, so any climate model has to adequately account for its behaviour.While many aspects of how the water molecule absorbs and emits light in our atmosphere are becoming increasingly well understood, some remain scientifically challenging. For example, a thorough understanding of the actual shape of individual water absorption lines is crucial for correctly modelling absorption by water vapour. This is necessary not only for climatic reasons but also because many remote sensing experiments, both space and ground based, require a detailed understanding of this absorption so that the characteristic signatures of other, trace species can be recovered. As discussed by Ngo et al. [1], there is now incontrovertible evidence that the true line shape, at line centre and in the near wings, departs significantly from the Voigt profile that has traditionally been used to represent the change in water line profiles as a function of temperature and pressure. However, what one should use instead of Voigt profiles remains a matter for debate, which is informed by high-precision experiments such as those reported by Hodges et al. [2].The radiative impact of atmospheric water vapour (and its role in remote sensing) depends not only on the tens of thousands of individual absorption lines but also on the so-called water vapour continuum, which varies relatively smoothly with wavelength from visible to microwave wavelengths. The continuum has been most intensively studied in the mid-infrared 'window' (wavelengths from approx. 8 to 12 mm) but, outside of this region, its strength and temperature dependence, particularly in atmospheric conditions, are less well characterized. In addition, there has been a long-lasting and robust debate as to the underlying cause of the continuum, it probably being due to some combination of the far wings of individual spectral lines of the water monomer and absorption owing to water dimers and other bimolecular complexes (e.g. H 2 O-N 2 ). Many radiative transfer codes used in weather prediction and climate models use the semi-empirical representation of the water vapour continuum from the CKD (Clough-Kneizys-Davies) and MT_CKD (Mlawer-Tobin-Clough-Kneizys-Davies) family of models; Mlawer et al. [3] provide a detailed description of the newest version of this model and describe how One contribution of 17 to a Theo Murphy Meeting Issue 'Water in the gas phase'.