A surface urban energy (water) balance model (SUEWS/LUMPS) was driven by 1yr field observations to evaluate the model capability in Shanghai urban area. The model input parameters were partly localized by field survey. The results show that the model can well reproduce the diurnal pattern of radiation components, the daily peak of net allwave radiation flux (Q*) was underestimated by about 25 W·m-2， the amplitude of the diurnal variation of the downward longwave radiation flux (L↓) was underestimated for all seasons, and the performance for upward longwave radiation flux (L↑) was much better than L↓. The occurrence time of the daily peak of sensible heat flux (QH) for each season (except spring) was simulated well by SUEWS, while the magnitude of QH for each season was underestimated. The performance of SUEWS model for latent heat flux (QE) in summer and autumn was better than LUMPS, but they were very similar during winter and spring. SUEWS model successfully reproduced the transition time of the sign of the storage heat flux (ΔQS) in winter, spring and autumn, but had a lag about 2 h in summer. The seasonal variation of the model performance for ΔQS was relatively large. Analysis of the model error dependency on a set of meteorological variables indicates that there was an increased underestimation of QH and QE and increased overestimation of ΔQS under higher air temperature or stronger wind speed conditions. The QH had an obvious underestimation by about -50 W·m-2 for easterly wind while overestimation by about 15 W·m-2 for the westerly wind was mainly the dense buildings over the east side and the green park over the west side was not considered by the model.