1. O(2) intakes were determined on subjects running and walking at various constant speeds, (a) against wind of up to 18.5 m/sec (37 knots) in velocity, and (b) on gradients ranging from 2 to 8%.2. In running and walking against wind, O(2) intakes increased as the square of wind velocity.3. In running on gradients the relation of O(2) intake and lifting work was linear and independent of speed. In walking on gradients the relation was linear at work rates above 300 kg m/min, but curvilinear at lower work rates.4. In a 65 kg athlete running at 4.45 m/sec (marathon speed) V(O2) increased from 3.0 l./min with minimal wind to 5.0 l./min at a wind velocity of 18.5 m/sec. The corresponding values for a 75 kg subject walking at 1.25 m/sec were 0.8 l./min with minimal wind and 3.1 l./min at a wind velocity of 18.5 m/sec.5. Direct measurements of wind pressure on shapes of similar area to one of the subjects yielded higher values than those predicted from the relation of wind velocity and lifting work at equal O(2) intakes. Horizontal work against wind was more efficient than vertical work against gravity.6. The energy cost of overcoming air resistance in track running may be 7.5% of the total energy cost at middle distance speed and 13% at sprint speed. Running 1 m behind another runner virtually eliminated air resistance and reduced V(O2) by 6.5% at middle distance speed.