1. The blood flow through the forearm was measured 2 sec after single, brief isometric hand-grip contractions. The tension and duration of those contractions varied from 10 to 100% of the maximal voluntary contraction (m.v.c.) and from 2 to 12 sec, respectively. 2. The blood flow increased linearly with tension up to about 60% m.v.c. but further increases in tension, up to 100% m.v.c., did not elicit higher blood flows than were found at 60% m.v.c. The same relationship between tension and the resultant blood flow held for all durations of contractions, from 2 to 12 sec. The blood flow immediately after (2 sec) contractions at a given tension increased linearly with the duration of the contraction, from 2 to 12 sec. Maximal exercise blood flow was approached only in response to the longest contractions (12 sec) at tensions of 60% m.v.c. or higher. 3. Brief alterations (2--5 sec) of transmural pressure across blood vessels did not result in a significant change of blood flow, either in the resting forearm or when the vessels were dilated by brief, isometric contractions. When the tension was applied or released either rapidly or gradually ('ramp' contractions) there was no correlation between the rate of change of stretch on arterial vessels and the resultant blood flow. However, there was a direct relationship between a force--time integral (duration of contraction x peak tension) and blood flow. All these results make it clear that changes in blood flow in the forearm elicited by brief isometric contractions are not the result of a myogenic reflex but are metabolically induced. 4. Successive contractions exerted at 60% m.v.c. for 4 sec induced a blood flow of 21.2 +/- 1.6 ml.min-1.100 ml.-1 when a rest interval of 8 sec was allowed between the contractions. Blood flows remained constant at this submaximal level, even when muscular fatigue was induced, and when there was an accompanying large increase in blood pressure. 5. Isometric muscular activity by the contralateral arm which resulted in fatigue, associated with a large increase in mean blood pressure, did not alter the level of vasodilation that was induced by brief, isometric contractions in the 'test' arm. 6. It is suggested that the vasodilatation in response to intermittent isometric contractions is the result of metabolic vasodilatation of distal segments and continued sympathetic vasoconstriction of the proximal segments of the forearm vascular bed.