Mindfulness skills are a component of many modern cognitive-behavioral therapies that are used to treat a wide range of disorders, including psychotic disorders. While habitual (i.e., trait) mindfulness is associated with clinical outcomes, the effects of momentary (i.e., state) mindfulness are unclear. This is due in part to previous studies using cross-sectional designs relying on trait self-report questionnaires. Although such approaches are invaluable, they lack temporal specificity to evaluate momentary changes and effects of mindfulness. To address these limitations, the current study used ecological momentary assessment (EMA) to evaluate state levels of two mindfulness skills, acceptance and monitoring, and their association with state fluctuations in symptoms. Participants included individuals with affective and non-affective psychotic disorders (PD; n = 49) and healthy controls (CN; n = 53) who completed six days of EMA. Results indicated that the PD group endorsed lower state acceptance than CN; however, the groups did not significantly differ in monitoring. Further, greater state mindfulness skills in both acceptance and monitoring were associated with greater positive affect, reduced negative affect, and reduced negative symptoms. However, participants with a predominantly affective psychosis presentation showed differential effects compared to those with non-affective presentations. These findings suggest that mindfulness training for people with psychotic disorders may benefit from focusing on improving acceptance in order to improve emotional experience and build on existing monitoring skills. Further, mindfulness based psychosocial interventions may offer a novel means of treating negative symptoms in people with PD, which are currently stalled and largely unresponsive to other treatments.