Mental disorders are common in all countries, aff ect every community and age group, contribute substantially to the overall burden of disease, and have major economic and social consequences and eff ects on human rights. However, the greatest inequities are cross-national: 80% of people aff ected by mental disorders live in low-income and middle-income countries, which benefi t from scarcely 10% of global mental health resources. Global mental health initiatives attempt to improve the availability of, access to, and quality of services for people with mental disorders worldwide. Diagnostic categories and a classifi cation of mental disorders, which are essential to achieve objectives of global mental health, are needed for a range of stakeholders: for health-care practitioners to make treatment decisions and implement clinical guidelines; for policy makers to make decisions about allocation of resources; and for patients and their families to gain an understanding of their disorders. But can contemporary psychiatric classifi cations meet these needs? Non-specialist health professionals working in routine health-care settings deliver more than 90% of mental health care worldwide. Psychiatric disorders are frequently diagnosed in epidemiological surveys in community and primary care populations, in particular the common mental disorders of depression, anxiety, somatoform, and stress-related disorders (all of which are distinct categories in contemporary classifi cations). However, a large gap often exists between the numbers reported in surveys and those recorded by primary care workers. The response of psychiatry has traditionally been to assume that these workers are not well informed about contemporary psychiatric concepts. Many strategies, such as the use of simple diagnostic algorithms, straightforward management guidelines, training programmes to improve skills, and advocacy campaigns, were developed and implemented to address this gap. However, these approaches did not seem to substantially change primary care worker behaviour, or improve detection rates or outcomes in developed nations or low and middle-income countries. Indeed, one of us (VP) had proposed that even the term “mental” needed to be dropped altogether from such training. The American Psychiatric Association, who produced the Diagnostic and Statistical Manual of Mental Disorders fourth edition (DSM-IV), and WHO, who produced the International Classifi cation of Diseases tenth revision (ICD-10), also published concise versions of these classifi cations for use in primary care (DSM-IV PC and the ICD-10 PHC). They reduced the number of categories (eg, from hundreds of ICD-10 categories to 26 in ICD-10 PHC), the many categories of depressive and anxiety disorders were reduced into single categories, and the more obscure diagnostic categories (eg, dissociative stupor) were dropped altogether. However, these simplifi ed schemes remained virtually unknown and unused in primary care worldwide. Nevertheless, there are plans to release similar primary care versions (DSM-5 PC and ICD-11 PHC) to correspond to the latest revisions of these classifi cations. Given the allegiance that these classifi cations owe to their parent classifi cations, we doubt that they will be suitable for the global mental health cause. The vast diff erences in settings, patient populations, and perspectives between psychiatrists and primary health-care professionals demand caution in the translation of specialist concepts and classifi cations for use in primary care. Primary care workers typically see patients with mild and non-specifi c symptoms, subsyndromal and mixed presentations, often clustered around the case threshold and frequently associated with psychosocial adversity and with physical health problems. They recognise the importance of psychosocial circumstances (eg, stress, personal resources, coping, social supports, and culture) and their eff ect on mental health. They prefer not to use mental disorder labels because of the high rates of spontaneous remission and placebo response and the absence of improvement with antidepressant drugs in those with mild disorders. A related serious concern for primary care workers is the medicalisation of human distress. They contend that the use of symptoms to diagnose mental disorders, without consideration of context, in particular psychosocial adversity, essentially fl ags nonclinically signifi cant distress, especially at lower degrees of severity. Although psychiatrists prefer to use disorder labels, primary care workers often favour dimensions of distress for presentations of common mental disorders. Primary care workers seem to be uncomfortable with the use of the notion of mental disorder, with its disease halo, which sidesteps the disease–illness dichotomy while attempting to encompass both disease and distress. Consequently, these practitioners mostly do not use psychiatric categories at all, preferring to avoid potentially stigmatising and meaningless labels, or use categories like “mixed anxiety–depression” and “adjustment disorders”. However, mixed anxiety and depression is not included in DSM-5, ICD-11-PHC, or in the recent WHO guidelines, which are designed explicitly for primary care. Adjustment disorders are also excluded from ICD-11-PHC and the WHO Mental Lancet 2014; 383: 1433–35