Changes in solar activity are regularly forwarded as an hypothesis to explain the observed global warming over the last century. The support of such claims is largely statistical, as knowledge of the physical relationships is limited. The statistical evidence is revisited. Changing solar activity is a statistically plausible hypothesis for the observed warming, if short-term natural variability is the only alternative explanation. Compared to the enhanced greenhouse effect, the solar hypothesis looses a substantial part of its plausibility. Reversely, the size and significance of the estimated impact of the enhanced greenhouse effect on the global mean temperature is hardly affected by solar activity.