Although cognitive theories of depression posit a causal role for information processing, popular treatments of depression have mostly focused on altering dysfunctional cognitive content, which may explain limited treatment effects. New experimental paradigms allow to modify underlying cognitive processes (cognitive bias, cognitive control). Both research lines have developed independently, and mostly remain disconnected from research focusing on the role of dysfunctional cognitive content. Relying on state-of-the-art statistical techniques, this project fills these gaps. WP1 combines intensive daily diary methods with two-step multilevel Vector Autoregressive models to establish and test the structure of the cognitive model and its temporal dynamics in high ruminators. Rooted in control theory, WP2 includes a simulation study to identify the optimal intervention strategy based on network dynamics, and puts this to the test using single-case series in remitted depressed (RMD) individuals, focusing on the temporal unfolding of effects. WP3 provides an experimental test of these relationships in a large-scale clinical trial, where RMD individuals will complete intervention modules targeting cognitive distortions and information processing, allowing to disentangle effects of key components of the model. The project concludes with a theoretical paper providing a quantification of the integrative cognitive model of depression, with direct implications for depression prevention.