Blood pressure was measured both directly and indirectly in seven hypertensive patients before and after a single oral dose of 100 mg metoprolol, at rest and during sustained handgrip. 2. Intra-arterially measured systolic and diastolic blood pressure increased linearly with time during sustained handgrip at 50% of maximal voluntary contraction. This linearity persisted for 60 s or more in most cases. Heart rate increased linearly for the first 30 s. 3. Indirectly measured blood pressure using an observer bias minimizing Auto-Manometer, under-read systolic and over-read diastolic pressure both at rest and during handgrip. 4. By exact timing of recorded values during handgrip and linear extrapolation (or interpolation) from base-line readings, mean values at 30 and 60 s of handgrip were calculated. The relationship between direct and indirect values remained the same at base-line and 30 s of handgrip. At 60 s of handgrip, this was true only for diastolic pressure. For systolic pressure, indirect and direct values almost coincided. 5. After metoprolol, directly recorded pressure fell slightly (7--12 mm Hg, 0.02 less than P less than 0.10), both at rest and during handgrip, and heart rate fell by 15--18 beats/min (P less than 0.01). The systolic blood pressure and heart rate effect of metoprolol at 1 min handgrip correlated with peak plasma drug levels. Indirectly measured blood pressure did not change significantly. 6. The rate of rise in heart rate and blood pressure from base-line to 60 s handgrip was not significantly influenced by metoprolol.