BACKGROUND: Uncertain prognosis presents a challenge for therapists in determining the most efficient course of rehabilitation treatment for individual patients. Cortical Sensorimotor network connectivity may have prognostic utility for upper extremity motor improvement because the integrity of the communication within the sensorimotor network forms the basis for neuroplasticity and recovery. OBJECTIVE: To investigate if pre-intervention sensorimotor connectivity predicts post-stroke upper extremity motor improvement following therapy. METHODS: Secondary analysis of a pilot triple-blind randomized controlled trial. Twelve chronic stroke survivors underwent 2-week task-practice therapy, while receiving vibratory stimulation for the treatment group and no stimulation for the control group. EEG connectivity was obtained pre-intervention. Motor improvement was quantified as change in the Box and Block Test from pre to post-therapy. The association between ipsilesional sensorimotor connectivity and motor improvement was examined using regression, controlling for group. For negative control, contralesional/interhemispheric connectivity and conventional predictors (initial clinical motor score, age, time post-stroke, lesion volume) were examined. RESULTS: Greater ipsilesional sensorimotor alpha connectivity was associated with greater upper extremity motor improvement following therapy for both groups (p < 0.05). Other factors were not significant. CONCLUSION: EEG connectivity may have a prognostic utility for individual patients’ upper extremity motor improvement following therapy in chronic stroke.