The skeleton consists of a series of elements with a variety of functions. In locations such as the skull, where shape or protection are of prime importance, the bone's architecture is achieved during growth under predominant genetic control. In locations such as the limbs, where the ability to withstand repetitive loading is important, only the general form of the bone will be achieved as a result of growth alone, the remaining characteristics resulting from an adaptive response to functional load bearing. In the horse, this functional load-bearing will be provided by the animal's natural activity pattern in box or paddock supplemented by the specific activities of the training regimen. It is the adaptive response to the total activity pattern that influences bone modelling and remodelling and so determines the bone's architecture. The objective of the training regimen is to ensure that this response achieves an appropriate match between bone architecture and the loads it is required to withstand during both training and athletic performance. We propose that for the match between architecture and load-bearing to be established, and maintained, subsequently, bone cells must be able to 'assess' directly or indirectly the functional strains produced within the bone tissue. Because these strains are used as feedback they are both the objective of functionally adaptive modelling and remodelling, and the stimulus for its control. The mechanisms whereby bone cells control skeletal modelling and remodelling to produce a functionally competent skeleton are unknown although some of the factors and sequence of events involved are presented here. The extent to which variation in training regimen affects bone architecture has not been studied systematically in any species.