In this study, lipofuscin was examined in the eyestalk ganglia of tagged European lobsters Homarus gammarus released into the wild at Scapa Flow, Orkney. Scotland, at approximately 3 mo of age and recaptured at ages between 5.4 and 9.6 yr. Lipofuscin deposits were often most abundant in cell cluster A of the medulla terminalis (MT-A), where they exhibited typical autofluorescence, histochemical, distributional and structural properties. Confocal fluorescence microscopy and image a'nalysis were used to quantify the deposits The study demonstrated that lipofuscin accumulation in the MT-A is age-dependent (r = 0 640, p -0.0002). For the available sample range, no other statistically significant relationships were found (carapace length vs age: r = 0.147, p = 0.359, body weight vs age: r = -0.054, p = 0.738; carapace length vs lipofuscin concentration: r = 0.331, p = 0.0849; body weight vs lipofuscin concentration: r = 0.181, p = 0.358). Body size had no age discriminating power. There was no difference in lipofuscin accumulation rate between males and females. When placed in perspective against the highest lipofuscin concentrations so far measured for wild individuals of this species, it was apparent that the available sample of tagged lobsters, spanning an age interval of about 4 yr, represented only a very small window of the total lifespan. The results indicate that measurement of MT-A lipofusc~n concentration will provide considerably more accurate age determination of fished lobsters than the current body-size-based approach. MT-A lipofuscin concentration correctly ages approximately 4 3 % of lobsters in a sample to within 1 yr or less of their true age, and 95 % of lobsters to within 3.5 yr Carapace length correctly ages only 3% of individuals correctly to withi.n 1 yr The width of the 95% confidence intervals for carapace-length-based age estimates are so large as to render these estimates meaningless. Size-at-age and number-at-age data obtained using lipofuscin will prove useful for estimating growth and mortality in wild lobster populations and provide insight into potential biases in the current, conventionally derived estimates of these parameters. From a practical perspective, it is easier to sample nerve tissue containing lipofuscin from the eyestalk than the brain and has the advantage of leaving the lobsters in marketable condition.