A recent paper in Geophysical Research Letters, 'Solar influence on winter severity in central Europe', by Sirocko et al (2012 Geophys. Res. Lett. 39 L16704) claims that 'weak solar activity is empirically related to extremely cold winter conditions in Europe' based on analyses of documentary evidence of freezing of the River Rhine in Germany and of the Reanalysis of the Twentieth Century (20C). However, our attempt to reproduce these findings failed. The documentary data appear to be selected subjectively and agree neither with instrumental observations nor with two other reconstructions based on documentary data. None of these datasets show significant connection between solar activity and winter severity in Europe beyond a common trend. The analysis of Sirocko et al of the 20C circulation and temperature is inconsistent with their time series analysis. A physically-motivated consistent methodology again fails to support the reported conclusions. We conclude that multiple lines of evidence contradict the findings of Sirocko et al.