The design and positioning of the fishway’s entrance is one of the crucial construction criteria, determining its overall effectiveness. In accordance with the literature, entrances are positioned as close as possible to the lateral barriers. However, specifically in this location, turbine outlets or weir discharges increase flow velocity, turbulence, particle suspension and noise, severely limiting high resolution underwater observations and in situ research.  Therefore, little is known of fish behaviour right in front of the entrance. Entrance strategies (e.g. do fish cross or avoid high velocity areas?), search patterns (e.g. at what distance or depth are dead ends surpassable?) and even abortion parameters (e.g. number, time, direction) are important for the entrance design. Especially without knowledge of the total amount of fish that are motivated for upstream migration, fishway efficiency assertions are incomplete. Sonar and multi beam sonars prove to be valuable instruments, able to deal with limited visibility but, at the moment, still display one main deficiency: only two dimensional data is obtained. In an experimental setup, we have overcome this deficiency by using two multi beam sonars perpendicular to each other and programmed an algorithm to combine the fish tracks from both 2D-data outputs into a 3D dataset. Besides pointing out important methodical pitfalls and challenges, we would like to present results providing new insights to entrance approach behaviour, the significance of river bed connection, vertical dead ends and overall detectability of fishway entrances.