Complex motor skills can be acquired while observing a model without physical practice. Transcranial direct-current stimulation (tDCS) applied to the primary motor cortex (M1) also facilitates motor learning. However, the effectiveness of observational practice for bimanual coordination skills is debated and there is little research on the effects of tDCS on acquiring bimanual skills and the underlying effective/causal brain connectivity. We compared the effect of primary motor cortex tDCS (M1-tDCS) to action-observation (AO) when acquiring a bimanual, two-ball juggling skill and characterized the brain causal connectivity patterns underlying each condition. Twenty healthy young adults with no juggling experience were randomly assigned to either video observation of a skilled juggler or anodal M1-tDCS. Thirty trials of juggling were performed and scored after the intervention. Resting-state EEG data were collected before and after the intervention. Information flow rate was applied to EEG source data to measure causal connectivity. Juggling scores were significantly higher in the AO group (p=.03). We found the strongest information exchange from (L) parietal to (R) parietal regions, strong bidirectional information exchange between (R) parietal and (R) occipital regions and an extensive network of activity that was (L) lateralized in the AO condition. In contrast, the M1-tDCS condition was characterized by bilateral long-range connections with the strongest information exchange from the (R) occipital region to the (R) temporal and (L) occipital regions. This study provides new results about the distinct network dynamics of priming the brain for skill acquisition using direct stimulation or indirect stimulation via action observation.