The accuracy of the TM-2420 ambulatory blood pressure monitor (A&D Co, Japan) has been assessed by the indirect method according to the recommendations of the Association for the Advancement of Medical Instrumentation (AAMI). Ninety subjects (43 men and 47 women), aged 19-89 with a range of systolic blood pressures (SBP) of 81-211 mmHg and diastolic blood pressures (DBP) of 32-113 mmHg were studied. The monitor was compared with two observers using the Hawksley random zero sphygomomanometer. The standard deviation of the difference (SDD) between the observers was 3.0 mmHg for SBP and 2.3 mmHg for DBP. The mean differences between observers were -0.31 mmHg for SBP and 0.32 mmHg for DBP (both NS). The average of three readings for each subject recorded by the monitor was compared with the average of the simultaneous readings by the observers. The SDD was 7.2 mmHg for SBP and 5.5 mmHg for DBP (within the recommendations for accuracy). The differences between methods were -0.98 mmHg for SBP and 0.18 mmHg for DBP; these differences were not significant (NS). The monitor was also assessed against direct intra-brachial artery pressure in 12 subjects (36 readings). The mean difference between the monitor and simultaneous individual intra-arterial reading was -9.5 mmHg for SBP (P less than 0.001) and 3.7 mmHg for DBP (P less than 0.001). The SDD's between methods were 12 mmHg for SBP and 5.0 mmHg for DBP. Use of the monitor in general clinical practice in 100 patients was also assessed. The rate of errors was low (6.8%), and the device found to be acceptable.(ABSTRACT TRUNCATED AT 250 WORDS)