The reliable estimation of the urban heat island intensity (UHII) within an urban canopy will benefit a number of fields of research such as urban design, public health, and building energy performance. A diagnostic equation for estimating the daily maximum UHII (UHIImax), which is based on routine meteorological data and basic urban properties, was extensively tested with the long-term observations from the metropolis of Nanjing, China. The field measurement has been carried out based on the Local Climate Zone (LCZ) scheme, which is a standardized classification protocol for urban temperature studies. The temperature data set of 658 days collected from six different LCZ classes was used for testing. The results show that the diagnostic equation produced a robust and satisfactory estimation of the daily UHIImax under various weather conditions. The overall performance of the equation for Nanjing, that has a root mean squared error (RMSE) of 0.95 degrees C and a median absolute error (MEAE) of 0.60 degrees C, closely approximates its performance for the 14 cities in northwestern Europe (RMSE = 0.91 degrees C, MEAE = 0.58 degrees C). The results also imply that the equation has the potential to be applicable to other cities. Another attempt in this study was to develop a method to apply the diagnostic equation to building energy simulations (BES). The hourly heat island intensities were derived from the estimated UHIImax based on the typical diurnal pattern of UHII evolution. An apartment was modeled with the BES program EnergyPlus to examine the validity of the proposed method. Using the modeled and measured temperature data as the inputs in the EnergyPlus simulations, the annual, monthly, and hourly energy demands for cooling and heating were investigated and compared. For all the six LCZ sites except the large low-rise case (LCZ 8), the impacts of the heat island effects at the neighborhood scale on building energy performance were reasonably evaluated with the method. The limitations and perspectives of the equation and its application to BES have also been discussed. (C) 2019 Elsevier B.V. All rights reserved.