Summary Background As treatments for constipation become increasingly available, it is important to know when to progress along the treatment algorithm if the patient is not better. Aim To establish the definition of failure of a treatment to provide adequate relief (F‐PAR) to support this management and referral process in patients with chronic constipation. Methods We conducted an international Delphi Survey among gastroenterologists and general practitioners with a special interest in chronic constipation. An initial questionnaire based on recognised rating scales was developed following a focus group. Data were collected from two subsequent rounds of questionnaires completed by all authors. Likert scales were used to establish a consensus on a shorter list of more severe symptoms. Results The initial focus group yielded a first round questionnaire with 84 statements. There was good consensus on symptom severity and a clear severity response curve, allowing 67 of the symptom‐severity pairings to be eliminated. Subsequently, a clear consensus was established on further reduction to eight symptom statements in the final definition, condensed by the steering committee into five diagnostic statements (after replicate statements had been removed). Conclusions We present an international consensus on chronic constipation, of five symptoms and their severities, any of which would be sufficient to provide clinical evidence of treatment failure. We also provide data representing an expert calibration of commonly used rating scales, thus allowing results of clinical trials expressed in terms of those scales to be converted into estimates of rates of provision of adequate relief.