Droplet　jetting　velocity　is　one　of　the　most　important　factors　affecting　the　quality　of　piezoelectric　ejection　printing.　Due　to　the　nonlinear　relationship　between　the　two,　predicting　the　droplet　jetting　velocity　by　conventional　methods　is　very　time-consuming　and　impractical.　We　propose　a　genetic　algorithm　(GA)　combined　with　a　back　propagation　neural　network　(BPNN)　to　predict　the　droplet　jetting　velocity.　The　network　topology　and　the　values　of　each　parameter　of　the　model　are　designed　and　validated　to　elucidate　the　relationship　between　the　input　parameters　of　the　piezoelectric　ejection　system　and　the　studied　droplet　ejection　velocity.　The　GA-BP　model　was　trained,　tested,　and　tuned　using　a　database　consisting　of　data　generated　from　finite　element　calculations,　and　the　results　showed　an　error　of　3.34%　between　the　predicted　and　finite　element　simulated　droplet　ejection　velocities,　which　demonstrates　the　reliability　and　robustness　of　the　method.　©　Published　under　licence　by　IOP　Publishing　Ltd.