This paper presents a methodology that combines airborne Light Detection and Ranging (LiDAR) data with medium-resolution optical Indian Remote Sensing (IRS) satellite imagery to predict stand canopy height in a study area in Aberfoyle, southern Scotland. Canopy height is an important forest variable that can provide information relating to forest structure, Yield Class, standing biomass and thereby associated carbon sequestration. LiDAR data were acquired in 2006 for an area of 20 km2, which were used to produce a canopy height model (CHM) that for the forest areas was validated against field plot measurements. In addition, a satellite image from the IRS sensor was acquired for the same year covering southern Scotland. The principal objective was to extend forest canopy height predictions from the LiDAR data over a larger area using the IRS satellite image scene. Leave-one-out cross-validation was used to compute the root mean square error (RMSE), relative RMSE (per cent) and associated mean deviation (bias) as a way of assessing the efficacy of the use of different groups of explanatory variables, i.e. satellite image bands. The lowest RMSE and associated mean deviation were obtained using Bands 1 and 4 from the satellite image as identified by the Variable Importance from the random forest algorithm. The k-Nearest Neighbour technique was implemented to predict canopy height at two scales: one for the areas with field measurements and the second for all forest areas within the LiDAR dataset. A good correlation was achieved between field measurements and the CHM, but worse results were obtained when the LiDAR data were combined with the satellite imagery