This　article　investigates　the　synchronization　of　delayed　reaction-diffusion　neural　networks　(RDNNs)　based　on　intermittent　sampled-data　(SD)　control　under　spatially　local　averaged　measurements　(SLAMs),　where　distributed　and　discrete　delays　are　introduced.　In　order　to　save　the　control　cost　and　reduce　communications　burden,　we　introduce　intermittent　SD　control　based　on　SLAMs　to　design　the　synchronization　scheme,　which　requires　only　a　limited　number　of　actuators　and　sensors.　Meanwhile,　we　construct　the　interval-dependent　Lyapunov　functional　(LF)　for　the　synchronization　error　system,　which　is　continuous　during　the　sampling　period.　Then,　according　to　the　designed　LF　and　several　inequality　techniques,　some　criteria　for　intermittent　SD　synchronization　are　established　to　ensure　that　the　error　system　is　exponentially　stable.　In　addition,　we　transform　the　controller　design　issue　into　a　feasibility　problem　of　liner　matrix　inequalities　(LMIs).　In　the　end,　two　numeric　examples　are　offered　to　ensure　the　feasibility　and　effectiveness　of　the　method.