1. This study used three techniques (bilateral phrenic nerve stimulation, motor cortex stimulation and quantitative electromyography) to assess the degree of activation of the diaphragm, intercostal-accessory muscles and abdominal muscles during postural tasks and respiratory manoeuvres. They included maximal inspiratory and expulsive efforts. 2. Bilateral phrenic nerve stimuli at supramaximal levels produced an average change in transdiaphragmatic pressure (Pdi) of 28 cmH2O during relaxation. During maximal inspiratory or expulsive efforts, all subjects were able to activate the diaphragm fully at functional residual capacity as judged by the failure of stimuli delivered during the voluntary efforts to increase Pdi. Peak voluntary Pdi was about 30% less for inspiratory than expulsive manoeuvres. 3. By contrast, transcranial activation of motor cortical output to the diaphragm and abdominal muscles produced an increment in abdominal pressure of 25 +/- 7 cmH2O during maximal voluntary expulsive efforts. Given the lack of response to phrenic nerve stimulation at similar voluntary pressures, this suggests that abdominal muscles, and not the diaphragm, fail to generate their full contractile force during maximal voluntary expulsive manoeuvres. 4. Motor cortical stimulation during weak inspiratory efforts produced a small reduction in oesophageal pressure (i.e. increase in net inspiratory force) of 7-14 cmH2O. This response could not be extinguished during maximal voluntary inspiratory efforts in two of three subjects. This occurred despite the cortical co-activation of 'antagonist' muscles in the chest wall and abdomen, and passive transmission of pressure from the abdominal to thoracic compartments. 5. Integrated electromyographic activity (EMG) recorded from abdominal muscles (rectus abdominis, external oblique) was greater during trunk flexion than during maximal expulsive efforts. Similarly, integrated EMG of the intercostal-accessory muscles (sternomastoid, scalenes, parasternal intercostals) was greater during tasks requiring head and/or neck flexion than during the maximal inspiratory efforts. 6. These data show that the diaphragm can be fully activated by the central nervous system during voluntary respiratory tasks but that other agonist 'respiratory' muscles need not be activated fully. Given the complex actions of 'inseries' respiratory muscles revealed here, it is argued that differences in the transdiaphragmatic pressure during various postural and respiratory tasks do not necessarily imply variation in the level of diaphragmatic neural drive.