Feeding and starvation experiments were carried out with Clione limacina sampled in Kongsfjorden (Svalbard, Arctic) during summer 2002. Dry mass and lipid mass, lipid class and fatty acid compositions were analysed. Specimens of C. limacina used for the feeding study had a mean length of 25 mm, a dry mass (DM) of 13.7 mg, and a moderate lipid content of 12.1%DM. Animals were allowed to ingest only one individual of its exclusive prey, Limacina helicina which had 8.0 mm in diameter, 21.4 mg DM and 8.7% lipid of ash-free DM. Five days after feeding, the dry mass of C. limacina had increased from 13.7 to 25.3 mg which corresponds to an uptake of about 80% of the ash-free DM (14.3 mg) of L. helicina. Lipid mass increased from 1.5 to 3.9 mg which is almost two times more the ingested lipid from L. helicina (1.2 mg lipid). Thus, the major portion of lipids was synthesised de novo by C. limacina from non-lipid compounds. These lipids were triacylglycerols (TAG) and 1-O-alkyldiacylglycerol ethers (DAGE), increasing from low proportions of 6.1% and 5.7% to 42.3% and 25.8%, respectively. Considerable de novo synthesis was observed for the monounsaturated fatty acids 16:1(n - 7), 17:1(n - 8), 18:1(n - 9), and 18:1(n - 7) and the alkyl moiety 16:0. The increase in the polyunsaturated fatty acids 22:6(n - 3), 20:5(n - 3), and 18:4(n - 3) corresponded with the amount available by ingestion of L. helicina, supporting that C. limacina is not able to synthesise polyunsaturates. After 15 days of digestion, dry mass and lipids dropped almost back to the initial values. During the 100-day starvation experiment, two groups of animals were separately considered as storage lipid-rich and lipid-poor animals because of their large differences in the amount and proportion of TAG and DAGE. Storage lipid-rich C. limacina were only found until day 50, whereas lipid-poor animals were present throughout the experiment. In the lipid-rich specimens, the levels of TAG were about twice that of DAGE. The proportions of TAG decreased considerably during the 50 days of starvation (from 48.3% to 25.1% of total lipid). DAGE, varying between 16.5% and 20.5%, showed only a small decrease. The lipid-poor animals survived 100 days of starvation, exhibiting low initial amounts and proportions of storage lipids which were nearly exhausted at the end. In all C. limacina specimens, the total lipid content remained almost constant showing that lipid and non-lipid components were simultaneously utilised. This implies that body shrinkage may be an important adaptation to long-term starvation. Based on these results, it is possible to estimate the potential survival period of lipid-rich C. limacina under food limitation. A model, which considers maturity and reproduction (egg production), reveals that lipid-rich specimens might be able to survive up to 260 days without food. © 2005 Elsevier B.V. All rights reserved.