The Freiburg Mindfulness Inventory (FMI) is a widely used questionnaire of self-reported mindfulness. However, doubts have been expressed as to whether an adequate comprehension of the items of the FMI is independent of one's mindfulness experience (ME). The aim of the present study was to determine with qualitative methods whether and how ME influences the response to the FMI items. Two groups, matched for gender, education, and age (N = 11 each), with and without mindfulness training, completed the FMI while at the same time applying the technique of thinking aloud. The protocols of the two samples were compared using three different strategies: (1) predefined criteria on the comprehension of each item developed by FMI experts, (2) a coding scheme developed to identify differences in specific cognitive processes, and (3) qualitative analysis of comprehension patterns. The results showed that (1) participants with ME fulfilled the item criteria for comprehension much more than participants without ME. (2) The coding scheme demonstrated greater comprehension difficulties in the sample without ME. Differences in judgment processes between groups could not be found. (3) Qualitative analysis revealed comprehension problem patterns especially for eight items for the comparison group. It is concluded that a modification of the wording of several FMI items is necessary and that there is insufficient construct validity to use the current FMI in mindfulness-naïve samples. This may also be true for other scales tapping into the assessment of the awareness component of mindfulness, and it is recommended to also check their construct validity.