Controlled experiments carried out over a three month period on lemon sole (Parophrys vetulus) show that tagging, using the Petersen type of tag, causes significantly more deaths among females than would normally occur, and it is suggested that this is true also for males. The rate of tagging mortality without competition from other types of mortality, extrapolated over a period of one year, was computed to be 41 and 49 per cent for males and females respectively. In neither sex did fish of different sizes die differentially as regards either the number of deaths or the rate at which they occurred. By affixing tags according to three degrees of looseness (0.76 mm., 1.52 mm., and 2.28 mm.) and to different positions on the fish (the nape, below the dorsal fin at the widest dorso-ventral position, and on the caudal peduncle) it was shown that the loosest attachment, and application below the dorsal fin, caused least injury to the fish, as well as the least number of deaths.