The aerosol first indirect effect is known to cool the Earth radiatively. However, its magnitude is very uncertain. One of the difficulties in deriving this effect is caused by the coherent variation between aerosol abundance and meteorological conditions. In this study, we demonstrate that evaluation of the aerosol first indirect effect based on comparisons of clouds with different aerosol concentrations suffers an influence of the different degrees of mixing between clean and polluted clouds. By introducing a new method capable to remove this influence, we show that the strength of the aerosol first indirect effect is about half of that estimated by many previous investigators.