Global mean sea level observations are necessary to answer the urgent questions about climate changes and their impact on socio-economy. At GeoForschungsZentrum/Geman Processing and Archiving Facility ERS altimeter data is used to systematically generate geophysical products such as sea surface topography, high-resolution geoid and short- and long-period sea surface height models. On the basis of this experience, fully reprocessed ERS-1 altimeter data is used to generated a time series of monthly sea surface height models from April 1992 to April 1995. The reprocessing consists of improved satellite ephemerides, merging of Grenoble tidal model, and application of range corrections due to timing errors. With the new data set the TOPEX/POSEIDON prelaunch accuracy requirements are fulfilled. The 3-year time series is taken to estimate the rate of change of global mean sea level. A careful treatment of seasonal effects is considered. A masking of continents, sea ice, and suspect sea surface heights is chosen that is common for all sea surface height models. The obtained rate of change is compared to external results from tide gauge records and TOPEX/POSEIDON data. The relation of sea level changes and sea surface temperature variations is examined by means of global monthly sea surface temperature maps. Both global wind speed and wave height maps are investigated and correlated with sea surface heights and sea surface temperatures in order to find other indicators of climate variations. The obtained rate of changes of the various global maps is compared to an atmospheric CO2 anomaly record, which is highly correlated to El Ni¿o events. The relatively short period of 3 years, however, does not allow definite conclusions with respect to possible long-term climate changes. ¿ 1998 American Geophysical Union