We assessed the effects of selective restriction of movements of the rib cage (Res,rc) and abdomen (Res,ab) on ventilatory pattern, transdiaphragmatic pressure (Pdi), and electrical activity of the diaphragm (Edi) in five normal subjects exercising at a constant work rate (80% of maximum power output) on a cycle ergometer till exhaustion. Restriction of movements was achieved by an inelastic corset applied tightly around the rib cage or abdomen. Edi was recorded by an esophageal electrode, rectified, and then integrated, and peak values during inspiration were measured. Each subject exercised at the same work rate on 3 days: with Res,rc, with Res,ab, and without restriction (control). Res,rc but not Res,ab reduced exercise time (tlim). Up to tlim, minute ventilation (VE) was similar in all three conditions. At any level of VE, however, Res,rc decreased tidal volume and inspiratory and expiratory time, whereas Res,ab had no effect on the pattern of breathing. Res,ab was associated with higher inspiratory Pdi swings at any level of VE, whereas peak Edi was similar to control. Inspiratory Pdi swings were the same with Res,rc as control, but the peak Edi for a given Pdi was greater with Res,rc (P less than 0.05). During Res,rc the abdominal pressure swings in expiration were greater than with Res,ab and control. We conclude that Res,rc altered the pattern of breathing in normal subjects in high-intensity exercise, decreased diaphragmatic contractility, increased abdominal muscle recruitment in expiration, and reduced tlim. On the other hand, Res,ab had no effect on breathing pattern or tlim but was associated with increased diaphragmatic contractility.