This study examines whether reactivity testing in a laboratory setting might identify individuals whose blood pressure (BP) fluctuates excessively outside of the laboratory. Eighty-eight normotensive and untreated hypertensive patients were studied. Patients' BPs were obtained repeatedly on two separate screening occasions. They were subsequently admitted to a research ward, and 4 days later their BPs were measured at resting baseline and in response to a series of stressful tasks. Blood pressure declined by 14/11 mm Hg +/- 1.2/1.1 mm Hg (SE) from screening to the in-patient resting baseline. The four stressful tasks all led to increases in BP over the in-patient resting baseline level. Individuals with increased amounts of task-induced reactivity in the laboratory were also more likely to be those whose BP declined the most from out-patient screening to hospitalized baseline. The average correlation between the drop in BP associated with hospitalization and the BP reactivity to each stressor task was r = 0.33 for systolic BP (p < 0.01) and r = 0.24 for diastolic BP (p < 0.05). Reactivity responses in a laboratory setting may convey information about blood pressure fluctuation in the real world.