Fast structural change in agriculture and other primary production has decreased employment opportunities and the number of enterprises in sparsely populated rural areas. Nevertheless, structural development has been only moderate in the ownership of rural assets, such as arable land, farm compounds and forests. Land ownership has even been fragmenting over the last decades, as an increasing number of people have maintained their capital in land and rural estates, when they have moved on to off-farm and urban occupations. Already about 750,000 hectares, i.e. 33% of the total arable land area, is owned in Finland by non-farming households and cultivated under lease contracts. The population of these non-farming and possibly urban landowners is already larger than the number of active farming households. The observed trends imply a large potential and need to design and implement new rural policies which increase the recreational value of rural assets, particularly for their urban owners, while supporting the development of sustainable and productive local economic activities. Nevertheless, a prerequisite for successful policies is a thorough knowledge about the landowner preferences. Currently we do not know, however, what the preferences, attitudes and values of the non-farming households are concerning their farmland and farm compounds. It is unknown how many of these owners already live in urban areas and what kinds of bridges their land ownership builds between the rural and urban communities in terms of rent sharing, income distribution, capital gains, investment expenditures, and demand for services. The research questions of the study are: 1) What are the non-market benefits of land ownership for non-farming landowners and how these benefits affect land-use decisions? 2) Which policy instruments are the most cost-efficient in promoting non-farming, urban landowners´ decisions to invest in land improvements and environmental programmes and to prevent land fragmentation? 3) How the land-use decisions of urban farmland owners are associated with the demand of rural services, the distribution of land rents and income and investment expenditures? The theoretical objectives of the project relate to applying utility maximisation model in landowners´ decision-making instead of the traditional profit maximisation model. In a utility maximisation framework, we can incorporate the non-market objectives of landowners in modelling the decision-making. The approach is to model preferences by random utility models and to estimate willingness to participate in rural and environmental programmes and land transactions by choice experiment methods (CE). The efficiency of empirical estimates is increased by explicitly augmenting the estimating equations by geographical and spatial information, such as local nature and zoning of the land and the permanent residency of its owners. The highly censored multinomial choice models are extended and developed towards the state of the art spatial econometric methods. The behavioural effects of conjectured policy means and their fiscal effects are simulated. Our results will provide new information for designing and implementing efficient rural policies, having a goal of maintaining rural livelihood in terms of sustainable local businesses, wealthy living environment and high recreational value for rural areas. The results will provide efficient and useful predictions on how landowners respond to market movements, policy programmes and even further, how this response depends on unobserved effects possibly due to location and geographical differences.