This short review summarizes a series of studies on the effects of expiratory flow limitation (EFL) at approximately 1 L/s during incremental exercise to maximal workload (Wmax) in normal subjects on exercise performance, respiratory muscle dynamics and control, and CO(2) elimination. Each subject served as his or her own control by performing the same protocol without EFL. Additionally, an index of cardiac output was measured before and after imposing EFL while the subjects exercised at Wmax, Wmax was reduced to 65% of control by severe dyspnoea. EFL forced a decrease in the shortening velocity of expiratory muscles, resulting in increased expiratory pressures which accounted for 66% of the variance in Borg scale ratings of dyspnoea. In spite of an increase in the shortening velocity of inspiratory muscles, inspiratory pressures and power increased, because EFL exercise induced hypercapnia, which increased the chemical drive to breathe. This was in part due to an increased alveolar dead space presumably resulting from a reduction in pulmonary capillary blood volume secondary to the high expiratory pressures. A vicious circle was established in which expiratory muscle pressures induced hypercapnia, which resulted in an even stronger expiratory muscle contraction. The imposition of EFL reduced cardiac output by 10% and decreased arterial O(2) saturation, reducing energy supplies to working locomotor and respiratory muscles. This model reproduces the most important clinical features of COPD, and these arise from ventilatory pump dysfunction rather than from the lung. It also leads to hypotheses that can be tested in patients with COPD.