Age composition is fundamental to understanding the population dynamics and productivity of a fish stock. The use of scales to estimate age can result in large errors inage data for long-lived species, usually due to compression of scale circuli with age. Patagonian toothfish (Dissostichus eleginoides) is considered to be a long-lived species whose age has been estimated using both scales and otoliths. We estimated the age of D. eleginoides caught off South Georgia usil~go toliths and scales. For scales, we made impressions on acetate slides; for otoliths, we used transverse sections prepared by baking and grinding the posterior and anterior sides. Using ANOVA, we compared data obtained from the two structures to test the hypothesis that otoliths and scales give the salvie age estimates, and compared the precisioli of age estimation for both structures and between readers.Ages estimated using scales were significantly less than those estimated using otoliths. For scales, bias occurred for both readers between readings; for otoliths, only one reader was biased. Residual variances indicated one reader was relatively more precise than the other in estimating age using otoliths, but less precise using scales. This reflected the comparative experience of the two readers in estimating the age of D. eleginoides using otoliths and scales.