A 1/6th water physical model of a 140 tons gas-stirred steel ladle is used to evaluate mixing times (τm at 95% of chemical uniformity) in a two phase system without slag (air-water) and in a more realistic three phase system (air-water-oil) to simulate the argon-steel-slag system and quantify the effect of the slag layer on the mixing time. Slag layer is kept constant at 0.004 m. Mixing times are estimated through measured changes in pH due to the addition of a tracer (NaOH 1 M). The effect of the following variables on the mixing time is evaluated for a single injector: gas flow rate (7, 17 y 37 l/min) and the injector position (R/r= 0, 1/3, ½, 2/3 and 4/5). Experimental results obtained in this work show good agreement when compared against mixing time correlations reported by Mazumdar for the two phase air-water case (no slag considered). Another comparison is done using the new concept called “effective bath height” proposed by Barati, where the mixing time is a function of the size of the slag layer since this layer dissipates part of the total amount of stirring energy introduced into the ladle by the injection of gas. Agreement is not good in this case. Finally, an estimation of the percentage of the stirring energy dissipated by the slag is computed, including other factors that govern the dissipation of stirring energy. Percentage of energy dissipated by the slag is found to be between 2.7 to 12 % depending on the process conditions.