Catch per unit effort (CPUE) data are presented from krill fishing operations in CCAMLR Subareas 48.1, 48.2, 48.3 (South Atlantic) for the period 1975/76 to 1987/88. Comparisons are carried out between subareas in single years, between years on a subarea and area basis, and between oceanic and shelf regions. Results show that the CPUE may differ between subareas, years and shelf/oceanic regions. Obviously, a consistent pattern of CPUE values does not occur over time for the entire area. High average CPUE were observed in the years 1975/76 and 1980/81, whereas the lowest CPUE occurred in Subarea 48.3 in 1977/78. The validity of the CPUE index is discussed as an indicator of krill abundance/biomass and caution is expressed in this regard. Depending on the circumstances, a commercial CPUE index may seriously under- or overestimate the amount of krill available in a statistical subarea. Overall length-frequency distributions from commercial fishing operations in general tend to show little interannual variation because several factors influence the representativeness of krill smaller than 35 mm in the catches.