The electromyographical (EMG) activity was studied during voluntary flexion movements of the trunk in erect standing man. The movements were performed at maximal velocity with successively increasing amplitude to cover the whole range of motion. The EMG activity was recorded from agonist-antagonist pairs of muscles at the ankle, knee, hip and trunk. The angular displacements at the corresponding joints were recorded using a Selspot optoelectronic system. The duration of initiating activity in prime movers (rectus abdominis and rectus femoris) as well as time to onset of activity in muscles braking the primary movement (erector spinae, gluteus maximus and hamstrings) were highly correlated with amplitude, duration, peak velocity and time to peak velocity of the movement (r = 0.59-0.91). The corresponding correlations for peak acceleration and deceleration of the movement were low (r = 0.03-0.38), indicating that acceleration and deceleration of a movement was not coded in the temporal aspects of the EMG. Onset of activity in rectus abdominis and rectus femoris as well as an early appearing burst of activity in vastus lateralis were invariant in relation to start of movement over the whole movement range. In the initial phase of a fast trunk flexion, activity in tibialis anterior appeared successively earlier with increasing movement amplitude. This resulted in a changed order of activation for the muscles from proximal to distal (rectus abdominis first) to distal to proximal (tibialis anterior first). Two different forms of associated postural adjustments are present during a fast trunk flexion, one early fast knee flexion and a later slower angle extension. Prior to knee flexion, no activity was recorded from muscles flexing at the knee implying that some other force must create a flexing torque around the knee. It is suggested that activity in rectus abdominis initiating the primary movement also initiates knee flexion through the upward pulling of pelvis. This would be possible since rectus femoris stabilizes the pelvis in relation to the leg, allowing the force in rectus abdominis to be transmitted below the hip joint and act extending around the ankle joint. However, when tibialis anterior is activated it stabilizes the shank which in turn will cause a knee flexion controlled by a lengthening contraction in vastus lateralis. During the subsequent ankle extension activity appears in lateral gastrocnemius and soleus causing the associated postural adjustment at the ankle. It can be concluded that activation of postural muscles prior to prime mover muscles is not always necessary.(ABSTRACT TRUNCATED AT 400 WORDS)