Background: Selenium is an essential trace element that also elicits toxic effects at modest intakes. Investigations of selenium metabolites in urine can help our understanding of the transformations taking place in the body that produce these beneficial and detrimental effects. There is, however, considerable discord in the scientific literature regarding the selenium metabolites thought to play important roles in these biotransformation processes. Approach: We critically assessed the published reports on selenium urinary metabolites, from the first report in 1969 to the present, in terms of the rigor of the data on which structures have been proposed. Content: We present and discuss data from ϳ60 publications reporting a total of 16 identified selenium metabolites in urine of humans or rats, a good model for human selenium metabolism. We assessed the analytical methods used and the validity of the ensuing structural assignments. Summary: Many of the studies of selenium metabolites in urine appear to have assigned incorrect structures to the compounds. The long-held view that trimethylselenonium ion is a major human urinary metabolite appears unjustified. On the other hand, recent work describing selenosugars as major urinary metabolites looks sound and provides a firm basis for future studies.