An integrative model of approach and avoidance goals in achievement contexts is proposed based on the concepts and principles of nonlinear dynamical systems. These goals are conceptualized as self-organizing systems in which information relevant to competence expectancies and perceptions of benefit and threat for the self is integrated with respect to either a single coherent state (static integration) or 2 conflicting coherent states (dynamic integration). Mathematical equations depict the emergence of approach and avoidance attractor dynamics from the proposed integration process. The model specifies the conditions under which states of goal involvement will display patterns of resistance (stability), oscillation between approach and avoidance (instability), or reversal (nonlinear change). As a result, this model of approach and avoidance goals equips previous dynamical models of self-regulation with greater social–cognitive substance. It also provides a parsimonious account of achievement motivation and resolves lingering inconsistencies and uncertainties in the achievement goal literature. Moreover, the dynamical systems approach, on which the model is based, provides a heuristic framework for generating new predictions regarding the ebb and flow of approach and avoidance in achievement contexts.