Recently, major known inconsistencies between observationally constrained and unconstrained modelbased estimates of the global mean direct aerosol radiative forcing (DARF) have been resolved1 . However, there is still considerable debate about the magnitude of DARF uncertainty2–8. A recurring question is whether current aerosol models adequately cover the full range of uncertainties6,9. There are several ways to calculate DARF from reported aerosol model results. Here we show that calculating DARF using a simple Monte Carlo technique, which accounts for the variance in relevant input parameters, provides an estimate of DARF with a significantly larger uncertainty range. The expanded range is close to observation-based uncertainty estimates. This suggests that an upward adjustment of modelled direct aerosol radiative forcing uncertainty may be needed to account for the limited number of models used in recent assessments. A doubling of the estimated uncertainty range for DARF, as described here, would have a substantial impact on the uncertainty in climate sensitivity10-12.