The aim of this work is the calculation of areal rainfall on the basis of runoff measurements. Based on the conceptual rainfall-runoff model COSERO, two different methods are presented to inversely simulate rainfall with runoff observations as input. To verify the existence, uniqueness and stability of the inverse rainfall, numerical experiments with synthetic hydrographs are carried out successfully. The application of the inverse model with runoff observations as driving input is performed for 2 catchments in the northern Austrian Alpine foothills. The results show that the inverse model is able to calculate areal rainfall in a high temporal resolution successfully. Compared to station observations, the inverse rainfall sums and time series have a similar goodness of fit, as the independent INCA rainfall analysis of Austrian Central Institute for Meteorology and Geodynamics (ZAMG). Compared to observations, the inverse rainfall estimates show larger rainfall intensities and a higher variance, due to oscillations in the inverse rainfall. It is however shown, that the oscillations can be eliminated with a simple weighted moving average filter. In addition, the influence of different methods for calculating potential evapotranspiration on the inverse rainfall is investigated. Due to the availability of high-resolution meteorological data (1x1 km, 15-/60-min) of global radiation, air temperature, wind speed and relative humidity, different calculations of potential evapotranspiration are performed and analyzed for Austria (84 000 km). Especially above 1500 m a.s.l., the potential evapotranspiration estimates calculated with the energy-balance based approach of ASCE Penman-Monteith are significantly higher, compared to the temperature-based methods of Hargreaves and Thornthwaite. It can be concluded that the usage of air temperature as a proxy is, especially at higher altitudes with lower air temperatures, insufficient to calculate potential evapotranspiration.