Objectives Past research of the Five Facet Mindfulness Questionnaire (FFMQ) lacks clear results regarding its factorial validity, item fitting, mindfulness in the general population, and on the higher order structure of mindfulness. We derived an alternative two‐factor higher order structure for the FFMQ, delineating the attentional and experiential aspects of mindfulness. Method Data of 640 persons from the Austrian community were used for primary analyses, and data of 333 Austrian students were used for cross‐validation. Confirmatory analyses and exploratory structural equation modeling (ESEM) were utilized to investigate psychometric and structural properties. Associations with related variables and indicators of mental health were examined. Results Confirmatory models fitted only poorly on the full 39‐item FFMQ. Fit was acceptable in an abridged 20‐item version in both samples. The Nonreact scale had only weak psychometric properties. ESEM analyses suggested a good fit of two higher order factors and revealed structural differences between the samples. Beneficial effects of mindfulness appeared to be uniquely associated with the experiential aspects of mindfulness. Strategies of emotion regulation showed differential associations with the two higher order factors in the two samples. Conclusions Our findings are relevant both with regard to conceptual issues on mindfulness and the assessment of mindfulness with the FFMQ. Replications in meditating samples and in patients are needed.