Although relative humidity (RH) sensors in measurements are known to be temperature-compensated by manufacturers, our field study shows that the RH observations from different solid-state RH sensors had measurable temperature and humidity dependencies. Our field observations were taken during 2002 and 2003 in Lincoln, NE. The RH bias due to temperature and humidity dependencies was considerable for atmospheric applications. Temperature-humidity-dependent transformation models for RH measurements are derived from the statistical analysis of simultaneous RH measurements from HMP45C, HMP233, MP101, HMP243, and a chilled mirror hygrometer. By using an independent set of observations to validate the RH bias models, it is shown that the data quality of RH observations can be significantly improved for RH sensors. The results include the temperature and humidity dependencies of RH biases for each sensor, the variations of RH bias magnitudes, and performance of RH bias transformation.