Synthesizing representative weather data sets that reflect both typical and extreme conditions is essential for reliable hygrothermal simulations, especially when considering future climates, which involve significant uncertainties. In this study, multiple representative weather data sets are generated using various methods, broadly categorized based on picking the representative months using variables reflecting the outdoor temperature, moisture or rain conditions. The performance of each method is evaluated by conducting hygrothermal simulations of a prefabricated wood-frame wall, using both the original long-term weather data and the synthesized representative data sets. A comparison of different outdoor weather variables and wall characteristics is carried out, examining both average and extreme conditions. The results indicate that the representative data sets based on outdoor dry-bulb temperature (TDY, ECY, and EWY) can effectively capture the characteristics and variability of the long-term dataset.