Research in climate change requires long term homogeneous climatological series with variations caused only by natural fluctuations of weather and climate. Inhomogeneous time series arise either due to abrupt discontinuities or to effects that gradually increase or decrease over time. Both can affect air temperature and precipitation. Discontinuities in climatic data are caused by changes in instrumentation, station location, time of observation, and the environment surrounding the station. Many of the above-mentioned changes are reported for the US Weather Service Network;A study was conducted to evaluate climatic time series for discontinuities. Besides maximum and minimum temperatures, derived variables included heat stress, growing degree days, growing season length, heating degree days, cooling degree days, last date in the spring with minimum temperature below 0°C, and first date in the fall with minimum temperature below 0°C;Daily records of maximum and minimum temperatures were used to generate the seven derived variables from 99 Iowa weather stations over the 1951-91 period. An interpolation procedure was chosen to generate the reference time series which used the 25 surrounding stations which were most correlated with the candidate station being analyzed for discontinuities. The maximum likelihood ratio test developed by Alexandersson (1986) was selected to test the significance and to estimate the magnitude of the discontinuity. Beside the statistical framework, station histories were inspected to match the result of the statistical test with the record of changes in station location, time of observation, and observer. Ninety percent of the stations examined possessed at least one discontinuity. The highest percentage of stations with discontinuities was observed for maximum temperature and heating degree day time series with 69 and 66 percent of stations showing discontinuities, respectively. Generally, but not always the discontinuity in maximum and/or minimum temperatures resulted in a discontinuity for derived variables. Many discontinuities could be associated with some change registered in the station history, but some could not. The statistical framework used in this study can be applied to check the relative homogeneity of an existing climatic data network.