Background: A significant share of meditation research relies on trait psychometric measures of mindfulness, which neglect the intricate and dynamic unfolding of experiential processes entailed by meditation over time. There have been, however, a few attempts to capture ongoing changes in experience during meditation. In this study, we integrate and expand upon three of these previous approaches to create a novel instrument, which we called the Lyon Assessment of Meditation Phenomenology (LAMP). This questionnaire encompasses contextual, conative, affective, somatic, attentional, cognitive, and meta-cognitive domains.Methods: Fifty-three experienced meditators completed it after each meditation session during an intensive 10-day retreat. We modeled the time courses of the individual answers to each question using generalized additive modeling, and automatically clustered participants using group-based trajectory modeling.Results: Over 60% of the assessed measures exhibited significant change during the retreat across the group, following distinct temporal trajectories. These trajectories reflected differences between meditation types (chiefly, focused attention and open monitoring) and individual expertise, supporting a previously proposed multi-dimensional phenomenological model of mindfulness. We also identified three clusters of individual temporal trajectories associated with prior meditative experience and difficulties experienced during the retreat. Finally, we replicated and extended core findings from mindfulness research on pain regulation.Conclusions: The proposed multidimensional experience-sampling approach provides a rich characterization of the dynamical aspects of meditative experience and may be usefully employed to deepen our knowledge of the phenomenology and mechanisms of meditation and meditation-based interventions.