Insomnia disorder is a subjective complaint of sleep dissatisfaction including both night-time and daytime symptoms. Currently there are three commonly used diagnostic manuals each with their own set of criteria, which is often credited for the wide range in insomnia prevalence reported by population-based studies, especially those with self-reported insomnia. However, there are limited studies directly comparing different criteria and little is known about associations with health outcomes. Thus, the aim of this study was to compare the most commonly used diagnostic criteria for insomnia from the literature and to explore the associations with a range of physical and mental health outcomes. We used data from 21,083 women and men from the seventh survey of the population-based Tromsø Study which included adults aged 40–99 years. A revised version of the Bergen Insomnia Scale was used to define insomnia based on the 4th (revised) and 5th edition of Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR and DSM5), the 10th edition of the International Classification of Diseases (ICD-10), and the 3rd edition of the International Classification of Sleep Disorders (ICSD-3). We found the following prevalence of insomnia: DSM-IV-TR 23.6 %, DSM5 8.5 %, ICD-10 9.9 % and ICSD-3 20.0 %. When looking at each symptom, we found over half the participants classified as having insomnia using the DSM-IV-TR and ICSD-3 criteria did not report having impaired daytime functioning at least three days per week. Overall, participants with DSM5 and ICD-10 insomnia appeared to have worse health profiles, based on a higher percentage meeting the cut-off for possible anxiety or depression, reporting a psychological problem or chronic pain, and using antidepressants, painkillers or sleeping pills. However logistic regression models showed largely the same health factors had the same association with the odds for being classified as having insomnia disorder from each set of criteria. Overall, this study suggests that insomnia prevalence may be overestimated if daytime symptoms are not adequately included in accordance with current guidelines.