Daydreaming, a common mental activity, can be excessive and accompanied by distress and impaired functioning in daily life. Although currently not formally identified by diagnostic manuals, daydreaming disorder (maladaptive daydreaming [MD]) is a clinically well-defined phenomenon. However, research is lacking regarding the diagnostic reliability of MD. Our aims were (a) to develop diagnostic criteria and a structured interview for MD, (b) to examine the reliability of this measure for distinguishing individuals with and without MD, and (c) to establish an optimal cutoff score for identifying clinical-level MD using an existing self-report measure. Thirty-one individuals who met screening criteria for MD and 31 matched controls completed the self-report measure and participated in 2 structured clinical interviews. Each participant was interviewed independently by 2 clinicians blind to the participant’s group membership. Cohen’s kappa values for the agreement rate between each interviewer and the screening criterion, and between the 2 interviewers, ranged from good to excellent (κ = .63–.84). A cutoff score of 50 on the self-report measure yielded nearly perfect sensitivity and specificity and good-to-excellent agreement between the self-report measure and the interview (κ = .68–.81). Our interviews were conducted over the Internet, rather than in person; results might have been influenced by self-selection; and interviewing wider samples is warranted. We found that MD can be diagnosed reliably using a structured interview developed for that purpose. The new diagnostic interview showed excellent agreement with a self-report measure for the disorder. Additionally, we identified a useful cutoff score for future self-report research.