AbstractTags are commonly used to uniquely identify fish in order to estimate population size, harvest rates, and fish behavior. However, some tags have the propensity to be shed (lost), which can bias results. To examine the shedding rates of external tags and the potential bias this introduces for common snook Centropomus undecimalis we marked adult snook with an internal 23‐mm half duplex passive integrated transponder (PIT) tag (as a permanent mark) and one of two external tags: a T‐bar anchor tag (44 mm; Floy, FD‐68B) or a dart tag (89 mm × 2 mm; Floy, FT‐1‐94). Fish were tagged along sandy beaches of the Gulf of Mexico in southwestern Florida during the spawning season in 2007–2009. Short‐term (37‐d) external tag retention success was 100% for both tags. However, long‐term (391‐d) retention success was low for external tags (T‐bar = 76%; dart = 38%). Given the limitations of and trade‐offs among tag types, choosing the most appropriate tag will continue to challenge researchers. Until more universal tags are developed, researchers must carefully consider the advantages and disadvantages of each tag type with respect to different project objectives.Received October 17, 2010; accepted May 19, 2011