During 1968-1971 a total of 494 flathead catfish were tagged with ring and/ or spaghetti tags. Of 175 recaptures, III were by the authors and 64 were by fishermen. Application of ring tags to the pectoral spine and spaghetti (T-bar, 1chor type) tags to the operculum is described. The anchor tags were applie t to the operculum by piercing the bone with the needle of the tag applicator. Ting tags of monel metal were held totally by encirclement of the base of the pectoral spine. Rates of tag loss were measured from fish marked and recaptured March 1968 through October 1971. For both tag types, the rate of tag loss was linear the first year. but exponential the second year after tagging. A discontinuous linear regression. separating the first and second years was used to calculate tag loss. The rate of tag loss the first year was 0.0665%/ day for ring tags and 0.0830%/ day for spaghetti tags; the second year. the rates were 0.1230% and 0.0895%/ day for ring and spaghetti tags respectively. Loss of ring tags was much higher the second year compared with the first year. Using both tags on each fish would give low probability for complete loss of identity (P=.6% in I()() days, =2.0% in one year). One fish, tagged with both tag types. was recaptured with both tags still firmly attached 1410 days (3.86 years) after tagging. Maximum tag retention times have been 1531 days (4.19 years) for a ring rag and 1448 days fora spaghetti tag.