Biomass produced from perennial energy crops, Miscanthus and willow or poplar grown as short-rotation coppice, is expected to contribute to UK renewable energy targets and reduce the carbon intensity of energy production. The UK Government has had incentives in place, targeting farmers and power plant investors to develop this market, but growth has been slower than anticipated. Market expansion requires farmers to select to grow these crops, and the construction of facilities, such as biomass power plants, to consume them. Farmer behaviour and preferences, including risk-aversion, are believed to be important to crop selection decisions. Existing research estimating the total potential resource has either only simplistically considered the farmer decision-making and opportunity costs, or has not considered spatial variability. No previous work has modelled the contingent interaction of farmers’ decisions with the construction of biomass facilities. This thesis provides an improved understanding of the behaviour of the perennial energy crop market in the UK, by addressing these limitations, to understand the spatial and temporal dynamics of energy crop adoption. It attempts to determine the factors that govern the rate and level of adoption, to quantify the greenhouse gas abatement potential, and to assess the cost effectiveness of policy mechanisms. A farm-scale mathematical programming model was implemented to represent the crop selection of a risk-averse farmer. This was applied using spatially specific data to produce maps and cost curves economic supply, for the UK. To represent the contingent interaction of supply and demand within the market, an agent-based model was then developed. The results indicate that perennial energy crop supply may be substantially lower than previously predicted, due to the time lags caused by the spatial diffusion of farmer adoption. The model shows time lags of 20 years, which is supported empirically by the analogue of oilseed rape adoption. Results from integrating a greenhouse gas emissions balance shows that directly supporting farmers, via establishment grants, can increase both the carbon equivalent emissions abatement potential and cost effectiveness of policy measure. Results also show a minimum cost of carbon abatement is produced from scenarios with an intermediate level of electricity generation subsidy. This suggests that there is a level of support for electricity generated from energy crops that reduces emissions in the most cost effective manner.