Proper sizing of energy systems in a zero-energy building is crucial to ensure that the zero-energy building can perform well during its operational process. Meteorological data should be well prepared as they are important inputs for the system design of a zero-energy building. Particularly, the extreme and typical weather data are crucial for the sizing of air-conditioning system and renewable system, respectively. However, the existing weather datasets including typical meteorological year (TMY) and extreme meteorological year (XMY) show limited extreme and typical weather information of multiple years, which would lead to improper system sizes and consequently result in deteriorated building performance. To fill this gap, this study develops a calibrated TMY weather file by applying a quantile mapping method on conventional TMY using multi-year weather data, which can accurately represent both extreme and typical weather information of multiple years, overcoming the limitations of the existing weather datasets. In the validated case, the calibrated TMY is developed using the weather data of 1979–2015 purchased from the Hong Kong Observatory. The results indicate that in comparison with the conventional TMY-based design, the calibrated TMY-based design substantially improves the building performance with reduction of 28 unmet hours, 11% improvement of load match ratio and 10.4% reduction of lifecycle cost. Therefore, the calibrated TMY shows great potentials to replace the conventional TMY for zero-energy building system design in practice. © 2022 Elsevier Ltd