Asthma differs from many other chronic conditions in that most key management decisions are made in non-specialist settings, such as general practitioner surgeries and accident and emergency departments. Diagnosis in primary care relies on recognition of a characteristic pattern of symptoms and the occurrence of asthma attacks, sometimes supplemented by basic lung function tests. Ongoing management is guided by the assessment of symptoms and simple lung function measures of airflow obstruction, with little attempt made to personalise management. This approach is flawed because the inadequate specificity of symptoms, as well as the low sensitivity of variable airflow obstruction, means that a diagnosis of asthma is often difficult to exclude with confidence. Moreover, even if diagnosed correctly, dissociation between inflammation, airflow obstruction, and symptoms means that a generalised stepwise approach to managing asthma on the basis of symptoms is unlikely to be successful in a substantial proportion of patients. As a result, effective treatments are used inefficiently, and outcomes are often worse than they could be. Rather than use of either a population-based or personalised approach for the diagnosis and management of asthma, we recommend a new combined approach, in which treatment decisions are driven by objective assessment of key treatable mechanistic traits.