Background <p>We assessed the feasibility and acceptability of conducting an epidemiological survey to estimate the distribution of mental disorders in a sample of adults with type 1 diabetes&#xa0;mellitus (T1D).</p> Methods <p>Eligible participants were adults with T1D recruited from four general practices in southeast England. Phase 1 included screening measures for mental disorders in the DSM-5. Participants at phase 1 were invited to a structured clinical interview for DSM-5 at phase 2. Feasibility parameters included the proportions of those identified as eligible, consenting, and completing either or both phases, and acceptability.</p> Results <p>The study population comprised 146 adults with T1D.&#xa0;72% (<i>n</i> = 105) had correct contact details, were eligible and invited. 52% (<i>n</i> = 55) completed phase 1, of which 45% (<i>n</i> = 25) completed phase 2. Some measures had high rates of missing values and three mental disorders had concordant phase 1–2 pairs, namely schizophrenia spectrum and other psychotic disorders, depressive disorders, and substance-related and addictive disorders.</p> Conclusions <p>Conducting a two-phase survey of mental disorders in people with T1D is feasible and acceptable and can be improved using methods to update current contact details; adding secondary care diabetes services (hospitals) as recruitment sites; reducing the screening measures; and omitting diagnostic interviews for those mental disorders already listed in routine medical records.</p>