Lindzen et al. (2001) suggested that tropical anvil clouds act as a strong negative feedback in the global climate system because the higher the surface temperature, the lower the cloud amount and the stronger the thermal emission to space. This study evaluates that hypothesis by using CERES TRMM satellite measurements over tropical oceans and perturbation analysis of the 3.5-box model used by Lindzen et al. The modeled radiative fluxes of Lindzen et al. are replaced by CERES directly observed broadband radiation fields. Perturbation analysis shows that model simulated climate feedback is determined by the radiative forcing differences of the anvil clouds and other climatological regions. CERES TRMM observations show that the anvil clouds have much higher albedos and moderately larger longwave fluxes than those assumed by Lindzen et al. As a result, for 22% reduction in the anvil clouds, the net radiative forcing obtained by the current study is small and positive. It changes from about 0.02 to 0.44 W/m**2 depending on the relationship between anvil clouds and other tropical regions, compared to the strong negative net radiative forcing of Lindzen et al. (varying from -1.7 to -3.9 W/m**2).