 This paper proposes a new method for estimating time series of inland water levels using satellite altimetry data. This method uses an extended outlier rejection algorithm combined with a Kalman filter to combine multiple missions altimeter data, including InVisit, ERS-2, JSON-1, JSON-2, Topex slash Poseidon, and Serol slash Altica. The resulting time series are compared to in-sit-to-data and other altimeter databases, showing RMS differences of up to 36 cm for lakes and 114 cm for rivers. This article was authored by C. Schwatka, D. Devmuring, W. Bosch, and others.