The plotting of cumulative sums (cusums), a technique of proven value in the detection of trends in data collected at intervals of time, may be modified to analyze circadian blood pressure patterns quantitatively. Mean 24-hour ambulatory blood pressure is taken as the reference value and is subtracted from each pressure value. The products of the remainders and the corresponding time intervals are summed in sequence and are plotted against time to form a modified cusum plot. The slope of the plot over any given time period equals the difference between mean blood pressure during that period and mean 24-hour blood pressure. Crest and trough blood pressures (the mean blood pressures of the 6-hour periods of highest and lowest pressures) may be identified as the 6-hour periods where plot slopes are most steeply ascending and descending, respectively. The magnitude of the circadian blood pressure change, defined as the difference between crest and trough blood pressure, is calculated from the difference between crest and trough plot slopes. The height of the cusum plot, which reflects pressure alteration extent and duration, may also be used as a measure of circadian pattern. The modified cusums technique and cusum-derived statistics are illustrated using ambulatory blood pressure profiles of hypothetical and actual hypertensive subjects. Independence from fixed time periods improves precision and reproducibility. Cusum-derived statistics are simply calculated from raw ambulatory data and should prove useful in the quantitative analysis of circadian blood pressure profiles.