Dilutions of intravenous medications may be inaccurate. The mixing technique may be a crucial factor.Three factors of dilution were tested: volume for dilution (large vs small), method for mixing (shaking vs inversion), and number of maneuvers (3 times vs 10).Dilutions of glucose in saline solution were made by nurses, after a random factorial plan. The judgment criteria were the comparison between measured (Cmes) and expected (Cexp) concentration.Cmes (n = 40) ranged from 89.5% to 123.6% of Cexp and was more accurate when made with a large volume (98.4% of Cexp vs 106.5%) and when mixed by inversion (100.6% of Cexp vs 104.6%).Inversion rather than shaking and dilution in a large volume is a simple procedure for bedside medication preparation that allows better accuracy. The 3 versus 10 mixing procedures resulted in the same accuracy, which may be important for these time-consuming procedures. These results should be confirmed in clinical situations.