Sensory input can remodel representations in the sensory cortex, and this effect is heavily influenced by attention to the stimulus. Here we ask whether pure sensory input can also influence the spatial distribution of sensory effects on motor cortical hand area (sensorimotor organization) and whether this is modulated by attention. Sensorimotor organization was tested by applying short (1.5 s) periods of low amplitude vibration to single intrinsic hand muscles and measuring motor cortex excitability with transcranial magnetic stimulation (TMS). In healthy subjects, sensorimotor organization in the hand is focal, with input from one hand muscle increasing motor-evoked potentials (MEPs), decreasing short and increasing long-interval intracortical inhibition (SICI and LICI) in the vibrated muscle ('homotopic' effects) and having opposite effects on neighbouring muscles ('heterotopic' effects). Here we show that a 15 min intervention of vibration applied simultaneously to two hand muscles can lead to long-term (> 30 min) changes in the spatial pattern of sensorimotor interaction. The amount and direction of the effects depended on the subject's attention during the intervention: if subjects attended to both muscles when they were receiving simultaneous vibration, subsequent short-term vibration applied to one of them produced 'homotopic' effects on both muscles. 'Heterotopic' effects on a muscle not vibrated during the intervention were unaffected. If subjects did not attend to simultaneous vibration, subsequent short-term vibration of the muscles involved in the intervention no longer had any effect on them although the 'heterotopic' effects on a muscle not involved in the intervention were unchanged. We conclude that a 15 min period of pure sensory input can remodel the way that subsequent sensory inputs interact with motor output, that the effects are specific for the motor output to muscles involved in the intervention and that they are modulated by the subject's attention.