Connectivity changes after brain lesions due to stroke are tightly linked to functional outcome. Recent analyses of fMRI time series indicate that dynamic functional network connectivity (dFNC), reflecting transient states of connectivity may capture network-level disruptions distant to the lesion site. Yet, the relevance of such dynamic connectivity patterns for motor recovery remains unclear. We, therefore, combined the analysis of static and dFNC and a repetitive transcranial magnetic stimulation (rTMS) lesion approach, to test whether dFNC provides region-specific insight into motor system reorganization after stroke. We focused on the contralesional primary motor cortex (M1) and anterior intraparietal sulcus (aIPS), two regions previously shown to modulate motor performance post-stroke in a time dependent manner. In 18 individuals in the chronic phase after stroke (with either persistent or recovered deficits) and 18 healthy participants, we analyzed static and dynamic resting-state connectivity. We then applied online rTMS intereference over contralesional aIPS and M1 during hand movement tasks to assess region-specific contributions to motor behavior. Consistent with previous studies, dFNC states were associated with persisting motor deficits, whereas static connectivity was not associated with motor outcome. dFNC but not static connectivity was associated with residual motor deficits and explained TMS-induced behavioral changes, when applying rTMS over contralesional M1. For contralesional aIPS, both static and dynamic connectivity were linked to TMS effects. This indicates that dFNC - more than static connectivity - contains information on the functional relevance of brain regions for motor outcome, specifically contralesional M1. Our results highlight the added value of temporal network analysis in understanding mechanisms of stroke recovery mechanisms.