Visualisation is now a fundamental component of e-science, allowing researchers to gain insight through visual representation of their data sets. Most Visualisation techniques have been developed on the assumption that the data values are exact - yet this is rarely the case. Typically data is uncertain - whether it be the result of measurement (where instruments introduce error) or simulation (where there are errors at all stages of the modelling and computational processes). There is increasing interest therefore in visualising not only the data but also the uncertainty associated with that data. The talk will look at current work in the field, and look in detail at recent work at Leeds by myself and my research student, Rodolfo Allendes, where we have tried to extend the traditional notions of contouring and surface plots to cover known uncertainty in the data. These new techniques will be illustrated with application to ocean data of interest to climate change predictions, in a collaboration with Keith Haines at University of Reading.