To compare the localized skin-cooling effects of 2 cryotherapy modalities and to review the clinical relevance of the results.Randomized controlled trial with repeated measures.Laboratory experiment.Convenience sample of 20 volunteers (13 women, 7 men), ages 21.3 to 44 years (mean, 31.3 +/- 6.8 y).A flexible frozen gel pack, a 454 g packet of frozen peas, or a control applied to the anterior thigh. No blinding was undertaken.Surface skin temperature under the modality at baseline and 10 and 20 minutes after application.Significant effects were recorded for modality (F(2) = 290.56, P <.0001), time (F(1.27) = 1868.07, P <.0001), and their interaction (F(2.09) = 305.47, P <.0001). After 20 minutes, frozen peas produced the lowest mean skin temperature +/- standard deviation of 10.8 degrees C +/- 2.28 degrees C compared with 14.4 degrees C +/- 2.53 degrees C from the gel pack and 26.1 degrees C +/- 1.75 degrees C from the control. Skin temperature fell between both time periods with the application of frozen peas but stabilized after 10 minutes of gel pack and control application.Application of frozen peas produced mean skin temperatures adequate to induce localized skin analgesia, to reduce nerve conduction velocity, and to reduce metabolic enzyme activity to clinically relevant levels. Flexible frozen gel packs did not cool skin sufficiently to achieve these levels.