Techniques to image surfaces of active stars from spectrometric data have been developed over the last decade, both for mapping brightness inhomogeneities (`Doppler imaging') and for magnetic fields (`Zeeman-Doppler imaging'; Brown et al. 1991; Donati 1992). Observations of a sequence of line profiles recorded at different phases of the rotation permits the reconstruction of a 2D brightness map, and in principle also of atmospheric density maps. Robust and efficient algorithms have been developed (see Vogt et al. 1987; Rice et al. 1989; Piskunov et al. 1990). Because of the long time coverage, the UV Spectrometer on STARS can provide a wealth of information on the structure of the upper atmospheres of cool, moderately active stars. Complete imaging is possible when ( Doppler line width). Important theoretical advances have recently been made on the formation of active regions (e.g. Choudhuri 1989; Moreno-Insertis et al. 1992). These have led to a prediction which STARS may verify, namely that G and K dwarfs with a rotation period smaller than 8-10 days must have high-latitude spots (Schüssler and Solanki 1992).