This thesis addresses the application of fusion as a powerful source of energetic neutrons. The conditions for maximising fusion power density, Pf, in a neutral beam fuelled plasma are determined and used to inform a high Pf device optimisation procedure. Restrictions on the externally applied power, not previously considered, are shown to be a crucial factor in determining the optimum plasma and reactor conditions. Two distinct regimes of operation exist, separated by a discontinuity in the optimum conditions. In one regime, beam-on-target reactions dominate and Pf is maximised when operating with a pure tritium target plasma. The optimum confinement is lower than that required for high gain operation but the energy multiplication is limited to Qf~1. Fast alpha particles, if confined, reduce Pf by a factor (1+0.2Qf)^-1, but leave Qf unchanged. In the second regime, thermonuclear reactions make a significant contribution to Pf, allowing for higher energy gain but requiring improved confinement. Alpha particle heating reduces Pf by around 20% but increases Qf. By considering established tokamak stability constraints the optimum plasma conditions are used to inform a high Pf device design. The model provides a framework for simplifying and informing what would otherwise be a perplexing search for the optimal reactor configuration and allows areas of particular interest to high Pf operation to be identified. A novel tokamak operating regime - the isothermal tokamak - is investigated. An analytic equilibrium is derived and the resulting density and current profiles found to be notably different from those of a conventional device. A model of the anomalous transport due to the trapped electron mode instability is derived. Simultaneous solutions to the MHD and transport equilibria are shown to only exist for relatively shallow density profiles.