The Outcome Questionnaire‐45 (OQ‐45; Lambert et al., ) has been designed for frequent assessment of a patient's functioning during the course of psychotherapy and has become one of the most frequently used outcome measures in the Netherlands. The OQ‐45 was originally normed on outpatients in secondary care only, but is applied in a wide variety of patient populations. As such, it has become increasingly important to provide cut‐off scores with the normal population, as well as between different patient populations. The present large‐scale Dutch study aims to provide cut‐off scores between several populations. Data were collected from the normal population (n = 1810) and patients in five different treatment settings: outpatient primary care (n = 1581), outpatient secondary care (n = 9433), private practice (n = 457), day patient (n = 481) and inpatient therapies (n = 485), a total of more than 14.000 administrations. Reliable change indices and cut‐off scores were calculated using the method of Jacobson and Truax (). The reliable change index for the patient population was calculated as 18 and the cut‐off between the normal and patient population as 56. Sensitivity, specificity and area under the curves of cut‐off scores between the normal population and the treatment settings were satisfactory and generally higher than 0.80. Between the patient populations, these measures were generally too low for strict use. The OQ‐45 total score can satisfactorily discriminate between the normal and patient populations. For assignment to specific treatment types, the OQ‐45 may help, but its use is somewhat limited in practice. Copyright © 2015 John Wiley & Sons, Ltd. Key Practitioner Message The Dutch OQ‐45 has satisfactory levels of reliability, sensitivity, specificity and area under the curve. The new overall cut‐off score for normal function for the Dutch OQ‐45 is 56 and the new reliable change index is 18. Cut‐off scores for several therapeutic treatments are provided.