The Devonian, like much of the Paleozoic, has long been a problematic period for paleomagnetism. Devonian paleomagnetic data are generally difficult to interpret and have complex partial or full overprints; problems that arise in data obtained from both sedimentary and igneous rocks. As a result, the reconstruction of tectonic plate motions, performed largely using apparent polar wander paths, has large uncertainty. Similarly, the Devonian geomagnetic polarity time scale is very poorly constrained. Paleointensity studies suggest that the field was much weaker than the modern field, and it has been hypothesised that this was accompanied by many polarity reversals (a hyperreversing field). We sampled middle to upper Devonian sections in Germany, Poland and Canada which show low conodont alteration indices, implying low thermal maturity. We show in this study that there are significant issues with these data, and they are not straightforward to interpret, even though no significant heating and remineralisation was likely to have caused overprinting. We compare our data to other magnetostratigraphic studies from the Devonian and review the polarity pattern as presented in the Geologic Time Scale. Combined with estimates for the strength of the magnetic field during the Devonian, we suggest that the field during the Devonian might have been so weak and in part non-dipolar that obtaining reliable primary paleomagnetic data from Devonian rocks is challenging. Careful examination of all data, no matter how unusual, is the best way to push forward our understanding of the Devonian magnetic field. Paleointensity studies show that the field during the Devonian had a similar low strength to that recently advocated during the Ediacaran. Independent evidence from malformed spores around the Devonian-Carboniferous boundary suggests that the terrestrial extinction connected to the Hangenberg event, was caused by increased UV-B radiation, supporting the weak field hypothesis. A fundamentally weak and possibly non-dipolar field during the Devonian could have been produced, in part, by true polar wander acting to maximise core-mantle heat flow in the equatorial region. It may also have influenced evolution and extinctions in this time period. There is a large-number of paleobiological crises in the Devonian, and we pose the question, did the Earth's magnetic field influence these crises?