AbstractTwo Italian homogenized data sets of sunshine duration (SD) and global radiation (Eg↓) relative anomalies are used to investigate to what extent these two variables agree with respect to their temporal evolution. They are compared for northern and southern Italy over the period 1959–2013. Both under all‐sky and clear‐sky conditions, the SD records tend to show a shorter and less intense decrease until the 1980s (“global dimming”) with respect to the Eg↓ ones, while there is a better agreement in the subsequent period when both variables increase (“brightening period”). To investigate whether such behavior can be explained by a different sensitivity of SD and Eg↓ to atmospheric turbidity variations, the observed clear‐sky trends are compared to those estimated by a model based both on Lambert‐Beer's law and on a simple estimation of diffuse radiation. Results show that most of the differences observed in the trends of the clear‐sky SD and Eg↓ records can be explained considering a realistic pattern of atmospheric turbidity in the 1959–2013 period. The only exception concerns winter and autumn in northern Italy where clear‐sky SD does not decrease in the dimming period as much as it would be expected on the basis of the corresponding increase in atmospheric turbidity. One reason for this discrepancy could be the influence of other variables like relative humidity. This case study highlights that changes in atmospheric turbidity have to be kept in mind when SD is used to investigate the multidecadal evolution of Eg↓.