We use data collected on the CCAMLR-IWC Krill Synoptic Survey (2000) to investigate relationships between cetacean density, krill density and oceanographic conditions. We explore the use of generalized additive models (GAMs) to model these relationships, and show how these models can provide abundance estimates for sub-areas within the survey region, as well as for the survey region itself. Abundance estimates from a simple conventional line transect analysis are also presented.