This　paper　performs　a　finite-time　analysis　(FTA)　and　implements　the　finite-time　anti-disturbance　synchronization　(FTADS)　control　for　general　switched　delayed　neural　networks　(SDNNs)　in　a　continuous-time　context.　While　the　prevailing　results　for　SDNNs　typically　mandate　the　finite-time　stability　of　individual　subnetworks,　this　investigation　explores　a　more　inclusive　scenario　in　which　all　subnetworks　of　SDNNs　may　exhibit　finite-time　instability.　To　address　this,　an　innovative　and　less　conservative　FTA　framework　is　established　by　developing　a　novel　time-dependent　multiple　Lyapunov-Krasovskii　functional　(TDMLF)　approach　and　incorporating　a　ranged　dwell　time　(RDT)　switching　strategy.　For　SDNNs　influenced　by　multiple　disturbances,　a　FTADS　criterion　is　established　within　the　proposed　finite-time　design　framework,　demonstrating　that　the　relevant　synchronization　error　can　achieve　the　desired　finite-time　boundedness.　In　the　end,　the　reliability　and　superiority　of　the　presented　FTA　approach　and　anti-disturbance　synchronization　control　technique　are　thoroughly　substantiated　via　two　demonstration　examples.