OBJECTIVES:  To determine the number of unique patient- and clinician-reported outcome (PRO) tools that have been used in studies published since 2005 for different mental disorders. METHODS:  We developed a PRO instrument ontology from those tools cited within the abstracts of over 100,000 studies identified by a systematic search of PubMed on the humanistic and economic burden of disease, which were stored in an online database (www.heoro.com). The ontology items are semantically indexed by general topic, population, disease or body area, symptom or treatment. Abstracts are tagged to each relevant ontology item during an automated process using natural language processing and tokenising approaches, and indexing is then checked by experts. RESULTS:  A total of 4,272 instruments were identified from 22,254 relevant abstracts. Of these, 2,540 (59%) were disease-specific tools, and 434 (17% of the disease-specific tools and 10% of the total) were used in studies of patients with mental disorders, more than any other disease area. Many of these tools (53) were not specific for any one mental disorder. The remaining 381 tools were designed for, or used in, more than 25 different mental disorders. The most diversity was found for affective disorders, with 44 tools for depression, 36 for anxiety, 16 for any affective disorder, 13 assessing mood, 8 for phobias and 7 for obsessive-compulsive disorder. Forty two tools were used for substance abuse or addictions, 41 each for dementia and stress/PTSD, 40 for schizophrenia, 16 for ADHD, 17 for eating disorders, 10 for delirium and 7 for bipolar disorder. CONCLUSIONS:  A vast number of different PRO tools have been developed and used in research published over the past 10 years. Mental disorders have been the most widely represented, reflecting their impact on subjective outcomes, but the diversity makes it challenging to compare effectiveness and cost-effectiveness outcomes across studies.