OBJECTIVE To determine the accuracy of 24-h ambulatory blood pressure monitoring, using the relationship between night-time and daytime values, in diagnosing secondary hypertension. METHODS A prospective study was performed in a referred population of 402 hypertensive patients (clinic systolic/diastolic blood pressure > 140/90 mmHg). The ambulatory monitoring data included 24-h mean, awake (daytime) and sleeping (night-time) values. Secondary hypertension was diagnosed by standard procedures. To describe the accuracy of ambulatory blood pressure monitoring, receiver-operator characteristic curves were constructed, using sensitivity and specificity values for deciles of the distribution of overnight blood pressure falls (absolute and percentage). Measurements included the fall in nocturnal blood pressure, sensitivity (the percentage of those with secondary hypertension who were classified as non-dippers), specificity (the percentage of non-secondary hypertensives who were classified as dippers) and predictive values of ambulatory blood pressure monitoring. RESULTS On average, overnight systolic/diastolic blood pressure fell in primary hypertensives (n = 290) by 20/18 mmHg (13%/19%), in white-coat hypertensives (n = 65, daytime ambulatory blood pressure <135/87 mmHg) by 17/15 mmHg (13%/19%) and in patients with secondary hypertension (n = 47, renal/renovascular and endocrine forms) by 13/11 mmHg (9%/12%). From receiver-operator characteristic curves, the nocturnal blood pressure fall of 15 mmHg showed the highest accuracy, with a sensitivity/specificity of 61%/69% (systolic) and 75%/62% (diastolic) whereas 10% (systolic) and 15% (diastolic) nocturnal falls had a sensitivity/specificity of 62%/74% (systolic) and 62%/70% (diastolic). The ambulatory blood pressure data had a high (>93%) negative predictive value for secondary hypertension. CONCLUSIONS Secondary hypertension is associated with a blunted nocturnal fall in blood pressure. Ambulatory blood pressure monitoring data are not critically important for the diagnosis and screening of secondary hypertension but may be helpful in excluding it.