Bubble curtains offer a potential solution for hindering downstream migrating fish from passing through the turbines of hydropower plants, by instead guiding them to a safe passage. Compared to the conventional method of using low-inclined bar racks close to the turbine intake, bubble curtains are considerably cheaper, especially for large scale hydropower plants. In this work bubble curtains are evaluated through an experimental parametric study, mainly regarding two quantities, defined here as the overall bubble coverage of the background, as well as the occurrence of large tears in the curtain. The examined parameters cover the design of the device that generates the bubbles, the positioning of the device, as well as the flow condition of the water. The bubble curtain was filmed with underwater cameras and image analysis was performed using image segmentation in MATLAB. Results indicate that the total air flow should be chosen high, around 200-500 nl/min/m. The number of holes the air is distributed across is of little importance, except in the area closest to the air outlets, where permanent tears may occur inbetween holes if the distance is to great. Smaller holes (<1mm) are preferable to larger holes (>1 mm), but the difference is small compared to the effect of the total air flow. In flowing water, a smaller angle of attack against the current is better based on the defined quantities, but it will also result in a longer and therefore more expensive device. The bubble curtain should not be placed where mean water velocities exceed 0,75 m/s, since it will result in an inclination with the current close to that of bar racks designed to guide fish towards the surface. Furthermore, at these water velocities some fish may have difficulties following the curtain, due to their limited swimming capabilities.