We investigate the extent to which aerosol extinction is a suitable proxy for the aerosol affecting drop formation. First we use multiple realizations of a cloud model to investigate the sensitivity of cloud drop effective radius re to aerosol parameters (size distribution and composition) and dynamical parameters (updraft and liquid water content). In general, re is most sensitive to cloud liquid water, a parameter often ignored in indirect effect analyses. The relative importance of the other parameters varies for different conditions but aerosol concentration Na is consistently important. Updraft plays an increasingly important role under high aerosol loadings. A breakdown of the individual aerosol terms contributing to drop size change shows that use of aerosol extinction as a proxy for size distribution and composition tends to underestimate the magnitude of the first indirect effect. This may influence interpretation of current satellite and surface remote measurements of the indirect effect.