Background Infusion of cold fluids can be used to induce hypothermia after cardiac arrest. Fluid temperature higher than 4°C could increase the volume needed, prolong the induction phase and/or contribute to complications. In this study, we analyzed the effect of flow rate and venous tubing insulation on the fluid temperature at the level of the intravenous (iv.) cannula, when infusing cold saline. We also analyzed the warming rate of 0.5 L bags of normal saline wrapped in ice packs and without ice packs. Methods All measurements were performed at ambient temperature 23°C, constant humidity of 40 % and without exposure to direct sunlight. 0.5L bags containing normal saline were stored in refrigerated conditions for at least 24 hours prior to usage, mean initial temperature in the bags was 3.96 ± 0.17°C. For measurements of temperature at the level of the venous cannula during an infusion, 0.5 L bags of normal saline were connected to venous tubing and iv. cannulas. 10 measurements of fluid temperature in the cannula and in the bag were performed (with one digital filament thermometers inserted in the bag and one in the cannula) at flow rates 10, 30, 60 and 100 ml/min. Measurements were repeated with just the venous tubing wrapped in aluminum foil for thermal insulation. 3 sets of measurements with 0.5 L bags of normal saline wrapped in ice packs and 3 sets without ice packs were made for the analysis of the rate of warming. Temperature of the fluid in the bag was recorded at 5 min intervals for 120 min with a digital filament thermometer inserted in the fluid. Results With non-insulated tubing, we observed significantly higher temperatures in the cannula compared to temperatures in the bag at all flow rates (all p<0.0001, differences between temperature in the cannula and temperature in the bag 8.8 ± 0.1°C, 4.8 ± 0.1°C, 4.1 ± 0.2°C and 3.0 ± 0.1°C at flow rates 10, 30, 60 and 100 ml/min, respectively). Temperature gain with insulated tubing was significantly lower compared to non-insulated tubing at all flow rates (p<0.0001), but also statistically significant at all flow rates (all p<0.0001, differences between temperature in the cannula and temperature in the bag 5.8 ± 0.1°C, 3.1 ± 0.2°C, 1.2 ± 0.4°C and 0.3 ± 0.1°C at flow rates 10, 30, 60 and 100 ml/min, respectively). Temperature differences between cannula and bag were also statistically different for different flow rates (all p<0.0001), for both insulated and non-insulated tubing. More specifically, the temperature differences at higher flow rates were significantly smaller for every pair of successive rates (i.e., 30 vs. 10, 60 vs. 30 and 100 vs. 60 ml/min). The mean rate of warming of bags not wrapped in ice packs was 6.9°C/h. The mean rate of warming of bags wrapped in ice packs was 3.4°C/h. The temperature of fluid in the bags wrapped in ice packs was significantly lower than in the not wrapped bags at every time point (from the time point at 5 minutes to 120 minutes, p<0.0001 for all time points). Conclusion When inducing mild hypothermia at ambient temperature 23°C we suggest using high flow rates (at least 100 ml/min) and heat insulated venous tubing to avoid excessive fluid temperature gain. Bags of cold fluid should be taken from the refrigerator just before starting the infusion and kept wrapped in ice packs to prevent warming.