Potential evapotranspiration (PE) is one of the required input time series to run most rainfall-runoff models, including the Sacramento Soil Moisture Accounting Model (SAC-SMA) in National Weather Service River Forecast System (NWSRFS). There are many different methods to compute PE, depending on what type of hydrometeorologic forcing data is available. This paper investigates several hydrometeorologic forcing data sets and tests and evaluates a number of PE schemes. The hydrometeorologic data sets to be examined include the NCDC Surface Airways observations, the University of Washington gridded meteorological data, the satellite solar radiation data from University of Maryland, and the NCAR/NCEP 50-year Reanalysis data. The PE schemes to be examined include the existing Synoptic Data Transfer (SYNTRAN) Program of NWSRFS, a temperature based PE scheme, and PE obtained from the NCEP LDAS NOAH model. The differences between the data sets and the PE estimates will be compared during selected common data periods. The long term differences (annual and monthly total) as well as short term differences (diurnal variation pattern and daily total) will be analyzed. The differences due to variations in climatic regimes will be studied as well. This study is motivated by the fact that traditional method of collecting sky cover information is being phased out with the full implementation of Advance Surface Observation System (ASOS) in the US. The ultimate goal of this study is to upgrade the SYNTRAN program in NWSRFS with enhanced capabilities to ingest new sources and new type of hydrometeorologic data.