The accuracy and clinical application of a new portable device for measuring ambulatory blood pressure (BP) (ABPM 630, Nippon Colin, Nagoya, Japan) were assessed. The device uses a conventional arm cuff inflated by CO2 gas from a compact cartridge and is based on a cuff-oscillometric as well as a Korotkoff sound (microphone) technique. Blood pressure values obtained by ABPM 630 were compared with those measured by the auscultatory method. With the microphone method the mean differences from the auscultatory method were -0.28 +/- 6.15 mm Hg (mean +/- SD) for SBP and 0.96 +/- 6.28 mm Hg for DBP (n = 256), while for the cuff-oscillometric method the mean differences were -1.77 +/- 6.07 mm Hg for SBP and 3.06 +/- 6.87 mm Hg for DBP (n = 297). There was a highly significant correlation between BP values measured by the auscultatory method and ABPM 630. In 40 untreated subjects, 24 h BP was monitored simultaneously with the ABPM 630 and with a finger volume-oscillometric device (UBP-100, UEDA, Tokyo, Japan). The daytime average of SBP with the former (126 +/- 11.6 mm Hg) was almost the same as that with the latter (123 +/- 16.0 mm Hg), while the nighttime average in the former (117 +/- 9.7 mm Hg) was significantly higher than that in the latter (108 +/- 14.1 mm Hg, P less than .01). Only 4 out of 40 subjects experienced no sleep disturbance from the arm-cuff inflation. Five of the 40 subjects complained that their sleep was frequently interrupted by the arm-cuff inflation.(ABSTRACT TRUNCATED AT 250 WORDS)