This research aims to address some of the limitations inherent in the methods currently used for identification of Housing Market Areas (HMAs) in Scotland. Firstly the conventionally defined geography of HMAs for the four largest cities in Scotland is evaluated using a spatially disaggregated series of individual house-move data for 2008. Some limitations with the conventional geography are revealed, and HMA regions are shown to be relatively inflexible in representing the strength of migration linkages within the market area, and take no account of the differences in the distances over which house moves take place in different Scottish cities. An alternative empirical approach is then explored, which takes advantage of improvements in the level of detail at which house moves are now geo-referenced in Scotland. Using an extensive dataset of sales transactions from the Registers of Scotland Executive Agency (RoS) for houses sold in 2008, migration distance is explored as a possible factor in determining HMA extent. Kernel Density Estimation is used to define and visualise various putative housing market areas in the different cities based on these disaggregated sales transactions. The resulting market areas are compared with those identified by the conventional methodology to allow the new methods to be evaluated. Although the results are based on data from a single year’s transactions, the utility and flexibility of developing a new approach to recognise HMAs using such disaggregated data and methods of analysis by GIS are indicated.