Many juvenile fish species are associated with structural habitats, potentially benefiting them from reduced predation and competition as well as enhanced feeding opportunities. It is also possible that structural habitats may provide a refuge from flow. Research on juvenile snapper in New Zealand has largely focused on their habitat preference. Juvenile snapper (Pagrus auratus) have been observed in close proximity to seagrass (Zostera muelleri) in New Zealand estuaries and I tested whether this association offered a refuge from flow. In an annular flume I exposed two cohorts of juvenile snapper 4.1 ± 0.5 cm and 8.9 ± 0.7 cm in length to sequential increases of flow speed (approximately 1 body length per second every 15 minutes). Flow speed increased at 3 cm s⁻¹ every 15 minutes for the smaller cohort of juvenile snapper, and 8 cm s⁻¹ for the larger cohort of juvenile snapper. Juvenile snapper were exposed to three treatments comprising of; bare, mixed (seagrass, bare and edge habitat) and full seagrass coverage. Fish behaviour was observed and critical swimming speed (Ucrit) estimated (i.e. the flow at which fish can no longer maintain position). Juvenile snapper were exposed to flow speeds that are representative of current flows in New Zealand estuaries and harbours. A startle response was recorded at the end as an indicator of fatigue and from larger juveniles a blood sample was collected via caudal puncture to determine if there were any physiological advantages offered by seagrass, by analysing stress indicators (based on lactate, triglyceride, glucose, haemoglobin and haematocrit). Velocity profiles indicated that the seagrass treatment and sampling points 2 (seagrass upstream and downstream) and 9 (5 cm into the leading edge of seagrass) from the mixed treatment dampened the flow speed within the annular flume. This result may be responsible for the increased percentage of time spent utilising the edge habitat as flow velocity increased, indicating that juvenile snapper seek refuge from high flows in or on the leading edge of seagrass patches. These results are consistent with an increase in critical swimming speed with increasing seagrass coverage (smaller juvenile snapper – bare = 19.47, mixed = 21.13, seagrass = 21.66 cm s⁻¹) (larger juvenile snapper – bare = 52.23, mixed = 58.89, seagrass = 60.31 cm s⁻¹). Physiological indicators, triglyceride, yielded significant differences between the bare treatment (1.40 mM) compared to the mixed (1.05 mM) and seagrass (1.01 mM) treatments, suggesting that energy stores were more readily utilised as energy expenditure increased where structural complexity was absent. The bare treatment also produced the highest values for both lactate and glucose, however, results were not significant. Mean cell haemoglobin concentration and total haemoglobin concentration did not yield significant differences between treatments. Whilst producing positive trends, results suggest that the effect of structural complexity was subtle in the annular flume. These findings have important implications for other hypotheses to explain the association between juvenile snapper and seagrass beds.