The usefulness of low temperature in-situ EPR spectroscopy to observe trapped holes and electrons in titania photocatalysts was first demonstrated by Howe and Graetzel. Since then, several groups have used the technique to observe charge separation in anatase-rutile mixtures, and to monitor reactivity of trapped electrons with oxygen, for example. The lifetimes of the holes and electrons, and their ease of reaction with adsorbed molecules, are crucial in determining the photocatalytic activity of titania and in other applications such as dye sensitised solar cells. In this work we are studying systematically the effects of surface hydroxylation, nanostructure and impurity levels on the generation and trapping of holes and electrons when titania is irradiated in vacuo at temperatures down to 4 K. We will describe how commercial nanocrystalline anatase samples have been found to contain nitrogen and or transition metal impurities, and how these participate directly in hole and electron trapping. Nanocrystalline rutile behaves quite differently from anatase in the in-situ EPR experiment.Trapped electrons are seen in rutile (as Ti3+) only after UV irradiation is stopped , implying the existence of delocalised conduction band electrons during steady state irradiation. The location of trap states within the band gap of rutile can be determined by selective wavelength bleaching experiments. We will report also on the influence of adsorbed species on the creation and trapping of holes and electrons in anatase and rutile.