BACKGROUND: In synchronized intermittent mandatory ventilation, it is generally accepted that the work of the inspiratory muscles is decreased by the ventilator so that their activity can be modulated by the frequency of assisted breaths. We examined the validity of this concept, which recently has been questioned. METHODS: We studied five patients receiving synchronized intermittent mandatory ventilation because of an acute exacerbation of chronic obstructive pulmonary disease. The level of machine assistance, defined as the percentage of total ventilation delivered by the ventilator, was varied from a high (> 60%) to a medium (20-50%) and to the lowest tolerated value (0% in four patients). Esophageal pressure, air flow, and the electromyograms of the diaphragm and sternocleidomastoid muscles were recorded. At each level of machine assistance, distinguishing assisted from spontaneous breaths, the duration of electrical activation, the integrated electromyograms, and the work of breathing were computed. RESULTS: The durations of electrical activation and the integrated electromyograms of the diaphragm and sternocleidomastoid were similar in successive spontaneous and assisted breaths. At > 60% of machine assistance, the cumulative values per minute of the integrated electromyograms of the diaphragm and sternocleidomastoid and the work of breathing were reduced only by 38, 32, and 44%, respectively, compared with the lowest tolerated level of machine assistance. The durations of electrical activation did not change with increasing level of machine assistance. CONCLUSIONS: The degree of inspiratory muscle rest achieved by synchronized intermittent mandatory ventilation is not proportional to the level of machine assistance; furthermore, the inspiratory motor output is not regulated breath by breath but rather is constant for a given level of machine assistance.