We examined patterns of habitat selection in wild-caught and hatchery-reared early juvenile red drum, Sciaenops ocellatus, using mesocosm experiments. Experiments were performed in the presence and absence of a pinfish, Lagodon rhomboides. We hypothesized that newly recruited red drum would have distinct patterns of habitat selection, and these patterns could be influenced by the presence of a pinfish predator. Red drum were introduced to replicate mesocosms containing all possible pair-wise comparisons of four different habitat types: marsh Spartina alterniflora, non-vegetated bottom/sand, oyster reef Crassostrea virginica, and seagrass Halodule wrightii. Wild-caught and hatchery-reared red drum showed distinctively different patterns of habitat selection. In general, wild-caught red drum selected structured habitats, while hatchery-reared fish did not show strong selection for any habitat type. When a predator was present, wild-caught red drum either changed habitat selection or showed significant selection for other structured habitats. This predator effect was similar albeit weaker for hatchery-reared red drum, but as in the trials without a predator, overall habitat selection was reduced compared to wild-caught red drum. Our results suggest that in the absence of seagrass, other habitat types such as marshes and oyster reefs may be important recruitment habitat for red drum. Additionally, hatchery-dependent behaviors may need to be assessed in designing stock enhancement programs.