When tested in isolation, stimuli associated with respiratory CO2 exchange, feedforward central command and type III-IV muscle afferent feedback have each been shown to be capable of eliciting exercise-like cardio-ventilatory responses, but their relative contributions in a setting of physiological exercise remains controversial. We reasoned that in order to determine whether any of these regulators are obligatory to the exercise hyperpnoea each needs to be removed or significantly diminished in a setting of physiological steady-state exercise, during which all recognized stimuli (and other potential modulators) are normally operative. In the past few years we and others have used intrathecal fentanyl, a μ-opiate receptor agonist, in humans to reduce the input from type III-IV opiate-sensitive muscle afferents. During various types of intensities and durations of exercise a sustained hypoventilation, as well as reduced systemic pressure and cardioacceleration, were consistently observed with this blockade. These data provide the basis for the hypothesis that type III-IV muscle afferents are obligatory to the hyperpnoea of mild to moderate intensity rhythmic, large muscle, steady-state exercise. We discuss the limitations of these studies, the reasons for their disagreement with previous negative findings, the nature of the muscle afferent feedback stimulus and the need for future investigations.