This study focused on identifying the sex of lake sturgeon by measuring the sex hormones estradiol and testosterone, and the phosphoprotein vitellogenin (Vtg) in blood plasma by radioimmunoassay and enzyme-linked immunosorbent assay, respectively, and evaluating these techniques as tools in lake sturgeon population management. Surveys of the St Clair River (SCR) lake sturgeon population have characterized it as rebounding by having steady or increasing recruitment since 1997. However, researchers have not been able to effectively determine the sex for most of the sturgeon they capture because few fish caught during surveys are releasing gametes. A total of 115 fish were sampled from May through June in 2004 and 2005 from the SCR, Michigan, USA. Of these, only four females and eight males were verified (i.e. they were releasing gametes at time of capture), resulting in very few fish with which to validate blood hormone and Vtg biomarkers of sex. Fifty-six percent of the fish were assigned a sex designation based on biomarker criteria. Correspondence between actual gonadal sex and biomarker-directed classification was good for the small subset of fish for which gonadal sex was definitively determined. Moreover, application of the steroid values in a predictive sex assignment model developed for white sturgeon misclassified only the same two fish that were misclassified with the steroid and Vtg biomarkers. The experimental results suggest a sex ratio of 1 : 2.7 (F:M), however more conclusive methods are needed to confirm this ratio because so few fish were available for sex validation. Of the 43 males, 14 were within the legal slot limit, 11 were smaller than 1067 mm total length (TL), and 18 were larger than 1270 mm TL. All 15 females were larger than 1270 mm TL, and thus protected by the slot limit criteria. Considering that lake sturgeon are threatened in Michigan, an advantage to using blood plasma assays was that fish were not harmed, and sample collection was quick, simple, and inexpensive. However, because a sufficiently large number of fish could not be validated for gonadal sex due to handling restrictions given the fish's protected status, assignment of sex is not based on a robust multi-variate model. An immediate alternative may be to use other non-invasive field methods (e.g. ultrasound, fiber-optic endoscope) to provide a more timely classification while establishing well-validated plasma hormone and Vtg-based predictive models for sex assignment of lake sturgeon. ?? 2009 Blackwell Verlag, Berlin.