This　study　is　purposed　to　understand　the　connectivity　of　cerebral　cortex,　especially　the　spatio-temporal　pattern　of　movement.　Functional　magnetic　resonance　imaging　during　finger　movement　was　used　to　extract　networks　connecting　correlated　cortex　areas.　Motor　cortex　networks　were　constructed　with　activation　areas　obtained　with　statistical　analysis,　and　linking　strength　was　obtained　by　calculating　the　correlation　coefficients　of　fMRI　time　series.　The　equivalent　non-motor　cortex　networks　were　constructed　with　certain　distance　rules.　Analysis　of　the　networks　shows　that　the　graphic　and　dynamical　measures　of　motor　cortex　networks　and　non-motor　cortex　networks　are　significantly　different,　the　motor　cortex　networks　are　more　compact　and　have　high　statistical　independence　and　integration　than　the　non-motor　cortex　networks.