Courtillot et al. (2007) use empirical correlations between geomagnetic records and climatic proxies to propose a connection between the Earth's magnetic field, solar activity and climate. In this Comment, we point out a number of approximations which undermine these correlations and connections.