Every year, very large quantities of powdered milk are produced worldwide (~5 billion tonnes). Optimization of the manufacturing is the key to ensure product yield and profitability. This can be achieved by optimizing product viscosity during the drying and atomization of the milk. During drying, water is removed to reduce the energy required in the atomization column. Good viscosity control leads to a better evaporation and therefore lower water content. During atomization, droplet size has a direct effect on the heat required for drying and therefore the energy required to produce powdered milk. One of the methods to optimize atomization is by controlling the viscosity of the fluid before spraying as viscosity has a direct effect on droplet size. The viscometer used during the trial (MIVI, Sofraser) is a viscometer at resonance frequency working at a high shear rate. The active part of the sensor is a vibrating rod held in oscillation at its resonance frequency where the amplitude of the movement varies according to the viscosity of the product. The viscometer was rated for 500 bar and 200°C, with a full-scale range of 200 cP. The trials were performed at a Dutch milk processing company where the sensor was installed on a flow through-cell before the atomizer. The equipment was tested on 3 different products at the following operating conditions: atomization pressure between 150 and 225 bars, temperature in the cell ~75°C and flow rate of 2,800 L/h. The viscometer was used to control the viscosity of the incoming milk from the dryers and ensure the milk was within the optimal viscosity range for atomization. The setpoint viscosity for the 3 different formulations was 20, 50, and 180 cP. The viscometer allowed maintaining viscosity within ±1 σ for 85 to 90% of points and a maximum difference of 3% between setpoint and mean measurement. Depending on product 1 σ is between 5 and 10% of setpoint. Higher casein to whey protein ratio gave more stable results, as well as higher carbohydrate to fat ratio. This will not be explained as it is not part of the study. Finally, during the trials, the use of a viscometer allowed 2% energy savings which represent about $15,000 to $20,000/year (based on $0.07/kWh).