A number of applications rely on the availability of high resolution TIR data, both in terms of spatial and temporal resolution, as well as in terms of spectral configuration. A consensus exists among the thermal community regarding the lack of such a kind of data, since only one operational sensor is currently providing multispectral TIR data at high spatial resolution (the case of ASTER, at 90m spatial resolution and a 16-day revisit time). In this paper we propose a minimum multispectral configuration to allow the retrieval of both land surface temperature (LST) and emissivity (ε) from thermal infra-red (TIR) data. For this purpose, we propose a minimum configuration based on three TIR bands at around 8.9, 10.6 and 12.0μm, which allows the application of two-channel (or split-window) algorithms as well as the application of a temperature and emissivity separation (TES) method (or a combination of both) for LST retrieval. Results presented in this paper are based on studies performed for different candidate missions, but most of the presented results can be also valid in other contexts.