Recovering from work is essential for maintaining occupational well-being, health, motivation, and performance, but recovery is often difficult to achieve. In this study, we evaluated and compared the effectiveness of two (parallel) interventions aimed at promoting recovery: one based on mindfulness and one involving applying cognitive–behavioral strategies. Both interventions were embedded in a measurement burst design, which allowed us to examine the mechanisms underlying change or intervention success. To explore mechanisms of change, we used the stressor-detachment model as a theoretical framework. We operationalized the interventions’ effects in three ways: as changes from pretest to posttest, as changes in daily states, and as changes in daily associations. To this end, we used intensive longitudinal data to examine the roles that daily negative activation plays in detachment and strain. In a randomized controlled trial (N = 393), we administered three assessments of traits: pretest, posttest (8 weeks later), and follow-up (3 months after the posttest). We also administered 2 work weeks of experience sampling questionnaires (preintervention and postintervention). Latent change models and Bayes factor equivalence tests revealed that both interventions substantially—and to a similar extent—increased detachment. Bayesian multilevel path models showed improvements in all state variables, including improvements in negative activation, and provided some evidence that mindfulness-based and cognitive–behavioral approaches might tackle different processes at the daily level. We discuss theoretical implications for the literature on recovery from work and specifically for the stressor-detachment model.