For small radicals in mobile liquids, the anisotropy of the g-value and hyperfine tensors is normally averaged out by molecular rotations. In the very strong magnetic fields (1-30 T) now in use for chemical studies, this may not be true: Larmor precession is so fast that even small differences due to differences in orientation can have a significant effect. The 'reversion' or reversal of direction of magnetic field effects on some radical pair reactions above 1 T may be due in part to this effect. The proposal is supported by numerical calculations.