This thesis estimates and analyses the values of forest recreation in Europe and considers the spatial and time aspects of valuing existing and new forest sites, given preferences for forest characteristics. The analysis of forest recreation values in Europe is conducted with a meta-analysis, which has not previously been done for Europe and includes exogenous variables on site characteristics, which is also new to meta-analysis. The estimation of total forest recreation values is carried out on state owned forests in a region in Denmark in 1977 and 1997 using a mixed logit specification of the random utility model (RUM) combined with Geographic Information System (GIS). A series of benefit transfers are conducted over time and space and validated against the 'true' values to ascertain the efficiency of transfers under different conditions. Validated benefit transfers over 20 years have not been attempted previously, primarily due to a lack of adequate data.