The researchers behind the study believe patients' true cardiovascular risk may be masked by this gap.
A team led by Dr Johanna Landgraf of the Good Samaritan Hospital in California analysed BP measurements for 337 patients, who were tested using both methods simultaneously.
Mean systolic BP readings were 1.9mmHg higher when measured by an auscultatory device (133.2mmHg) than by oscillometry (131.3mmHg). The disparity was greatest in patients over 65, where the difference was 2.6mmHg.
Nearly one in eight patients had systolic BP readings which varied by more than 10mmHg across the two methods.
The authors said the findings had important clinical implications because many hypertension treatment studies used auscultatory readings from mercury manometers.
'Physicians could incorrectly conclude their hypertensive patients had achieved the recommended BP goals, when the patients might not have reached these goals,' the researchers said.
Dr George Kassianos, a member of the British Hypertension Society's Blood Pressure Device Monitoring Group, said the number of BP readings GPs and nurses had to take made it 'next to impossible' not to use automated devices.
All devices, whether they were automated or manual, had their own shortcomings, he pointed out.
'If automated devices are reading lower by 2mmHg, I would prefer that much more than individual operators' bias with auscultatory devices where the individual operator tends to round up the BP figures to zero or five,' he said.
'I would recommend we use the automated devices in order to remove the operator bias.'
A list of devices approved by the British Hypertension Society appears on the organisation's website, he added.