The thermal protection provided by two helicopter passenger immersion suits was evaluated. Suit A was a standard 'dry' suit and suit B was a 'dry' suit with inherent insulation provided by inflation of the outer shell of the suit. During four hour immersions in water at 4°C with simulated rain, wind and waves, suit B provided significantly (p < 0.01) better protection against the long-term effects of immersion than suit A. The skin and core temperature of subjects fell at slower rates over the immersion period when they wore Suit B, they shivered less, had lower heart rates and were more comfortable in this suit. The problems of testing and selecting appropriate immersion suits are discussed and it is concluded that tests of immersion suits should be as realistic as possible and, when this is so, 'dry' suits with inherent insulation which is unaffected by leakage are likely to perform better in cold water than those without such insulation.