A composite recruitment index based on density data derived from acoustic surveys and on length-frequency data obtained from either target or station net hauls is described. This composite index is compared with an index where density and length data are derived from standard station net hauls only. Comparison of these indices showed substantial differences, due in part to bias arising from a large number of krill in just one of the station net hauls. Therefore, when the number of station net hauls is limited, bias is likely to be reduced if target net haul data can be included by using the composite recruitment index. This index is used to describe the population structure of krill around South Georgia during seven cruises between 1989 and 1998. Before 1994 the number of 1+ krill found around South Georgia was very low (proportion of 1+ krill <7%), while between 1994 and 1998 the number of 1+ krill found in the population increased with a very successful recruitment year due to krill spawned in 1994/95.