Numerical simulations without hydrological processes tend to overestimate the near-surface temperatures over urban areas. This is presumably due to underestimation of surface latent heat flux. To test this hypothesis, the existing single-layer urban canopy model (SLUCM) within the Weather Research and Forecasting Model is evaluated over Houston, Texas. Three simulations were conducted during 24-26 August 2000. The simulations include the use of the default "BULK" urban scheme, the SLUCM without hydrological processes, and the SLUCM with hydrological processes. The results show that the BULK scheme was least accurate, and it overestimated the near-surface temperatures and winds over the urban regions. In the presence of urban hydrological processes, the SLUCM underestimates these parameters. An analysis of the surface heat fluxes suggests that the error in the BULK scheme is due to a lack of moisture at the urban surface, whereas the error in the SLUCM with hydrological processes is due to increases in moisture at the urban surface. These results confirm earlier studies in which changes in near-surface temperature were primarily due to the changes in the turbulent (latent and sensible heat) fluxes in the presence of hydrological processes. The contribution from radiative flux was about one-third of that from turbulent flux. In the absence of hydrological processes, however, the results indicate that the changes in radiative flux contribute more to the near-surface temperature changes than the turbulent heat flux. The implications of these results are discussed.