It has widely been presumed that children between the ages of 4 and 6 years do not possess the cognitive capacity and experience to understand constructs such as ability and personal competence. Identifying the aetiology of such understanding is a primary objective of achievement motivation research considering the behavioural implications of these constructs, in relation, for example, to participation level (Roberts, Kleiber & Duda, 1981). Evidence was produced by the present research which challenged some assumptions which have been made about the limitations of children's conceptual understanding and behaviour in the physical domain. Previous research by Nicholls (1978) has indicated that in the academic domain, young children (4-7 years) do not differentiate between effort and ability as causes of outcome. He suggests that children only gradually differentiate between these two constructs to employ ability in their verbal explanations at a later age. When Nicholls' (1978) protocol for assessing this conceptual development was employed in the physical domain, using children aged between 4 and 13 years, current investigations revealed the same developmental trajectory. However, when effort cues were removed, children aged between 4 and 6 years used ability as a referent in relation to academic but not physical tasks. Contrary to previous assumptions, this suggests that young children may perceive effort and ability as discrete constructs and that some children are able to verbalise their beliefs about academic ability from an early age. Their beliefs, or their capacity to verbalise these beliefs, appears to be specific to the domain under consideration. This, and other findings which are described below, support current suggestions that behavioural indices are more appropriate than verbal, interview based methodologies for assessing young children's conceptual understanding. Behavioural measures do not require the child to verbalise their beliefs, a capacity which does appear to be limited, particularly in relation to physical tasks. Interpretational differences by individuals at different developmentalevels may also be ameliorated by the use of behavioural measures. Both Fogel & Thelen (1987) and Piaget have suggested that behavioural measures can be used to identify change. In the present research a behavioural measure was used to indicate the stability of different levels of effort and ability understanding. Attempts were made to relate stability of conceptual developmental stage to behavioural stability on a motor task, based on the theoretical proposals of, for example, Fogel & Thelen (1987). These authors suggest that developmental phenomena exhibit alternate periods of stability and instability which is mirrored in, and can be measured by, stability of the individual's behaviour during different developmental stages. Perhaps due to factors such as insufficient sensitivity of the method employed to detect these feasibly small shifts in stability, only limited evidence was produced to support the proposal that effort and ability understanding demonstrates alternate periods of stability and instability. However, this experiment did lead to further investigations of the behaviour of children between 4 and 6 years of age when they were offered rewards for successful performance attempts on a motor task. Findings contradicted previous assumptions made about the maturity of young children's achievement related behaviour. When fixed payoff rewards were used to increase the saliency of young children's performance outcomes their behaviour suggested an increased awareness of, and capacity ii to, assess their own competence level. Subsequent investigations which examined young children's task related behaviour when they were offered variable payoffs as rewards also revealed more advanced behaviour than would be suggested by previous research. In this context, young children could provide behavioural estimates of their perceived competence which were accurate and could select levels of task difficulty which offered them realistic levels of challenge. This behaviour suggested an understanding of the competence required to achieve success on tasks of varying degrees of difficulty and whether or not their own competence matched these required levels. These children also seemed to understand the incentive value, and their probability, of succeeding on different levels of task difficulty. It appears that, given certain circumstances, young children can: employ ability related explanations for performance outcomes; accurately assess their own competence, and appropriately use task related information to adopt personal levels of challenge which are compatible with their own level of task competence.