Hatchery-reared abalone outplanted into the wild at higher than ambient densities often experience very high levels of mortality as a result of density-dependent predation by natural predators. We conducted 2 experiments to assess the effectiveness of different methods of reducing predation on outplanted northern abalone (Haliotis kamtschatkana). During the first experiment, we removed small predators from brick-filled habitat cages that excluded large predators. The survival rates of juvenile abalone (mean shell length (SL), 22 mm) were greater in cages from which sea stars and other predators were removed periodically than in control cages. During the second experiment, we constructed 30 fenced 1 -m2 outplanting plots and assigned 10 randomly to be controls and 20 to be filled with 1 of 4 complex substrates (0.3-m layer of cobbles, 0.3-m layer of boulders, 0.9-m layer of cobbles, and 0.9-m layer of boulders). We then released 30 large juvenile abalone (mean SL, 51.5 mm) into each plot. The number of live outplanted abalone remaining in the complex substrate plots was significantly greater than in the control plots during the first 6 days. Surveys of a 4-m radius around each plot showed that emigration was lower from complex substrate plots than from control plots, and fewer shell fragments were found. The addition of complex substrate to plots provided crypsis from large predators but also resulted in significantly higher densities of small predators. Outplanting lower densities of abalone into larger plots of natural high-substrate complexity might attract fewer predatory sea stars and crabs, and thus result in higher survival rates.