Ischaemic stroke commonly causes hand weakness followed by a variable recovery of function. Transcranial Magnetic Stimulation (TMS) has previously uncovered abnormalities in both corticospinal (Catano et al. 1996; Traversa et al. 2000) and intracortical (Liepert et al. 2000; Manganotti et al. 2002) excitability in both hemispheres after stroke. Although the first month sees the greatest functional gains there is little longitudinal data addressing physiological changes in this period. Moreover, although some physiological measures have been related to function the changes in this relationship with time are unknown. We set out to investigate the evolution of corticospinal and intracortical excitability and to relate these to hand function. Ten patients with first ever ischaemic stroke were studied frequently during the first month using TMS and clinical assessments – patients were followed up at 3 and 6 months. In the affected hemisphere (AH) active motor thresholds were raised and MEP recruitment curve (RC) gradients were reduced during the first month (unpaired t tests, P=0.004 and 0.009, respectively). Active thresholds improved with time (1st to 3 months, P=0.046) but remained higher than normal at 6 months. Gradients in the UH were raised initially compared to the AH (paired t test, P=0.001) and showed a trend reduction from the 1st to 3 months (P=0.051) – UH gradients in the 1st month were more raised after subcortical than cortical stroke (P=0.003). Short interval (SICI) and long interval (LICI) intracortical inhibition were abnormally weak in the AH (P=0.032 and 0.048, respectively) and remained unchanged. In the AH, SICI was weaker after cortical than subcortical stroke (P=0.007). Correlations between measures of hand function and corticospinal excitability in the AH were strong during the acute period (active thresholds r=-0.67; RC gradients r=0.77) but weaker at 3 months (r=-0.47 and r=0.54, respectively). By contrast, similar correlations with measures of intracortical excitability were weak initially (UH SICI r=-0.55; UH LICI r=-0.48) and strong at 3 months (r=-0.69 and -0.74, respectively). We conclude that while patients rely initially upon what remains of the original corticospinal projection, time and motor practice allow networks in non-primary areas to become organised – intracortical disinhibition may allow continued access to these networks in the subacute period. The correlations between intracortical excitability and functional measures became weak by 6 months. This period coincides with the proliferation of peri-lesional terminal fields described after ischaemic infarcts in monkeys (Dancause et al. 2005). Thus structural changes between 3 and 6 months may reduce reliance on disinhibition for continued access to alternative motor networks. The data suggest that recovery from stroke is a dynamic and staged process.