Where sea urchin harvest has been so intense that populations have drastically regressed, concerns have arisen about the effectiveness of harvesting management. According to the theory of phase transition in shallow rocky reefs between vegetated and barren habitats, sea urchin recruitment, a key population structuring process, seems hampered by some stabilizing feedback despite an end to local human harvest of sea urchins. To shed a light on predation effects on sea urchin recruits, a 27-day field experiment was conducted using mega-predator exclusion cages (40x40x40 cm, 1 cm in mesh size) in barren and turf substrates. To facilitate this, 672 recruits (1.1 ± 0.02 cm in size) reared under control conditions were positioned in groups of 42 in each experimental unit (n = 4). Exclusion of mega-predators had a significant effect regardless the substrate, since a higher number of recruits was found under cages both in turf and barren. However, the results showed that in uncaged treatments the size of recruits that survived was larger in turf than in barren, as in the former substrate predation had reduced the abundance of the smallest recruits, highlighting that mega-predator presence affects differently the size of the recruits that had survived depending on the substrate. Overall, these results provide valuable information to address restocking actions of sea urchin populations in overharvested areas, where algal turfs are widespread, and assist studies on habitat stability mechanisms.