In human subjects sustained static contractions of the quadriceps femoris in one leg were performed with the same absolute and the same relative intensity before and after partial neuromuscular blockade with either decamethonium or tubocurarine which reduced strength to about 50% of the control value. During the contractions performed with the same absolute force, the magnitude of the cardiovascular responses (heart rate and blood pressure) was greater during neuromuscular blockade than during control contractions. During the contractions involving the same relative force the magnitude of the cardiovascular responses was almost the same with and without neuromuscular blockade. These findings were independent of the drug used. The metabolic part of the exercise pressor reflex was assessed by the application of an arterial cuff 1/2 min before cessation of exercise and for the following 3 min of rest. Although heart rate and blood pressure decreased after cessation of exercise, application of the tourniquet resulted in higher post-exercise values and this effect was seen both with and without neuromuscular blockade. Muscle biopsies from the subjects' m. vastus lateralis were analysed for fast- and slow-twitch fibre composition showing 27-66% slow-twitch fibres. No correlation was found between cardiovascular responses to static exercise, with or without neuromuscular blockade, and fibre type predominance. The results suggest that the involvement of fast- or slow-twitch muscle fibres does not play a dominant role in the cardiovascular responses to static exercise in man. Both central command and reflex neural mechanisms are of importance, and it appears that these two control mechanisms are redundant and that neural occlusion may be operative. However, when partial neuromuscular blockade induces a disproportion between an increase in central command and a constant or decreasing muscle tension and metabolism, the larger signal arising from central command determines the magnitude of the cardiovascular responses.