How is lime handled in kinematic and non-kinematic events from childhood to adulthood? Do subjects take one dimension or two dimensions into account to infer a third relationship? What occurs when the temporal information is presented in a hypothetical form (premises and conclusion)? A theoretical model based on empirical data from children, adolescents and adults is presented to describe the organization and development of relations across three dimensions in kinematics (i.e., relations between time, velocity and distance) and non-kinematics (relations between duration, beginning and ending times). The basic assumption of the model is that ‘’temporal dimensions” are combined pairwise (dyadic relations) within structurally stable and structurally unstable cognitive systems, before yielding to undecidability concept (last level of triadic relations). Some predications of the model for decidable and un-decidable time problems are presented. Data analysis shows that, even for adults, the concept of undecidability is constructed much later for non-kinematic events (i.e.. in problems concerned with duration and succession) than for kinematic events when the information is presented in a hypothetical form (written or oral statements).