Radiata pine sapwood and heartwood were dried with or without presteaming and then impregnated by vacuum treatment with water, toluidine blue and fluorescein. Sapwood uptake was 0.571 g/cm3 and was not affected by pre-steaming. As expected, the uptake by heartwood that had not been pre-steamed was very low. Pre-steaming increased liquid uptake from 0.113 g/cm3 to 0.438 g/cm3. When the uptake by pre-steamed heartwood from radial, tangential and transverse surfaces was compared, the greatest increase was from the radial surfaces, suggesting that pre-steaming of heartwood resulted in changes to the tangential liquid flow pathways. The liquid flow pathways in sapwood consisted ofaxial and radial resin canals, ray parenchyma cells in both fusiform and uniseriate rays. Penetration into tracheids was also observed. Without pre-steaming, there was limited liquid flow into heartwood, and this was generally confined to resin canals and ray parenchyma. Pre-steaming of heartwood increased penetration of dye into the resin canal network, presumably due to removal or redistribution of resin. Fluorescein was also evident in bordered pits between tracheids, suggesting that one of the ways that pre-steaming increased heartwood treatability was by altering the condition ofbordered pits to allow greater conduction. The combination of fluorescein dye and confocal microscopy was found to be a particularly effective way of visualising flow patterns, as it was possible to examine thick sections, which avoided microtome damage at the section surface. Examination of dry wood also minimised the possibility of dye redistribution.