INTRODUCTION Detailed building energy simulation tools such as DOE- 2 (LBL 1981) and EnergyPlus (Crawley et al. 2000) are commonly used to design sustainable buildings. These tools require hourly typical year weather files in order to estimate building energy use and building indoor comfort. Several procedures do exist to develop typical weather data using a single year of hourly data that are selected to represent the range of weather patterns that can be found in a multi-year data set (Keeble 1990). Several approaches have been utilized to develop and format a typical weather year for building energy analysis including the ASHRAE Test Reference Year or TRY (ASHRAE 1976)@ Typical Meteorological Year or TMY (Hall et al. 1978)@ the Weather Year for Energy Calculations (Crow 1981)@ TMY2 (Marion and Urban 1995)@ ASHRAE International Weather for Energy Calculations or IWEC (Thevenard and Brunger 2002)@ and more recently TMY3 (Wilcox and Marion 2008). Other selection approaches have been proposed (Hui 1996). Limited analyses have been reported to assess the impact of the selection criteria for generating the typical weather year on predicting the performance of building energy systems (Arigirou et al. 1999 and Massie and Kreider 2001). In particular@ Argiriou et al. (1999) tested several different TMY weather files generation procedures for Athens with 20 years (1977 to 1996) measured weather data. They considered several configurations of weighting factors and four methods to generate typical weather year including: the TMY method (Hall et al. 1978)@ a Danish method (Lund and Eidorff 1980)@ Festa-Ratto method (Festa and Ratto 1993)@ and 20-year average meteorological year. They developed weather data evaluation system based on building and solar systems. Specifically@ they utilized a simple solar water heating system@ a building@ a photovoltaic system@ and a large scale solar heating system with inter-seasonal storage@ and PV system. TRNSYS is used in the evaluation analysis (Anon 2000). A modified Festa-Ratto method was found to provide the best data set for Athens. Massie and Kreider (2001) estimated the discrepancies between TMY and TMY2s in predicting the performance of a photovoltaic system and a wind turbine. In this paper@ a series of sensitivity analyses is presented to assess the impact of the typical weather selection criteria on building energy analysis results. In particular@ the impacts on annual energy use predictions of weighting factors associated with various weather variables and of the length of historical data used in the selection procedure are evaluated throughout the paper. The analysis is carried out for 10 U.S. sites for which measured weather data for at least 30 years are reported. First@ a brief description of the prototypical office building used in the simulation analysis is provided. Then@ the results of the simulation analyses are presented and discussed.