A continuous-time analogue neural circuit which is capable of identifying the K largest of unknown finite value N distinct inputs, where , located in an unknown range is proposed. The circuit model is described by a state equation and by an output equation. A corresponding functional block diagram of the circuit is presented as N feed-forward hard-limiting neurons and two feedback neurons, which are used to determine the dynamic shift of inputs. The circuit combines such properties as high accuracy and speed, low hardware implementation complexity, and independency of initial conditions. Simulation example demonstrates that the circuit state variable trajectories are globally stable and globally convergent to the KWTA operation from each initial value.