Coral seeding is a technique used to restore reefs and promote adaptation, yet many factors influence coral survival. To investigate the impact of fish predation on seeding success we tracked the survival of ex-situ reared Acropora millepora spat on seeding devices at six sites in the Keppel Islands (Woppaburra sea country, inshore, GBR). Reefs were selected from take and no-take reserves, and two device designs with fish-exclusion features were tested against a control. After one year, >75 % of devices had live corals, and marine reserve status and device design influenced survival across sites. Reefs with fishing had double the survival (26%) compared to sites prohibiting fishing (14%), and devices with fish-exclusion features outperformed control devices at 67% of sites. Windward reefs with higher wave exposure and diverse fish communities were more conducive to coral survival over leeward reefs with high sedimentation. Sites with the low-lying macroalgae (Lobophora) were better suited for coral seeding over sites with erect macroalgae (Sargassum, Caulerpa, or Colpomenia). Our study demonstrates the pre-existing boundaries defined by marine reserve can inform the success of coral seeding, but the ecological drivers at each site may be stronger predictors of survival.