1. Surface EMG recordings were made from left and right homologous mus
cle pairs in healthy adults. During each recording session subjects we
re requested to maintain a weak isometric contraction of both the left
and right muscle. 2.. Cross-correlation analysis of the two multiunit
EMG recordings from each pair of muscles was performed. Central peaks
of short duration (mean durations, 11.3-13.0 ms) were seen in correlo
grams constructed from multiunit EMG recordings obtained from left and
right diaphragm, rectus abdominis and masseter muscles. No central pe
aks were seen in correlograms constructed from the multiunit EMG recor
dings from left and right upper limb muscles. 3. To investigate descen
ding pathways to the homologous muscle pairs, the dominant motor corte
x was stimulated using a focal magnetic brain stimulator whilst record
ing from homologous muscle pairs. 4. Following magnetic stimulation of
the dominant motor cortex, a response was recorded from both right an
d left diaphragm, rectus abdominis and masseter muscles. In contrast,
when recording from homologous upper limb muscles, a response was only
seen contralateral to the side of stimulation. 5. The finding of shor
t duration central peaks in the cross-correlograms constructed from mu
ltiunit recordings from left and right diaphragm, rectus abdominis and
masseter, suggests that muscles such as these, that are normally co-a
ctivated, share a common drive. The mechanism is discussed and it is a
rgued that the time course of the central correlogram peaks is consist
ent with the hypothesis that they could be produced by a common drive
that arises from activity in last-order branched presynaptic fibres al
though presynaptic synchronization of last-order inputs is also likely
to be involved. 6. The results of the magnetic stimulation experiment
s suggest that this common drive may involve the corticospinal tract.
7. We saw no evidence for a common drive to left and right homologous
muscle pairs that may be voluntarily co-activated but often act indepe
ndently.