The input current stimulating a neuron is not directly measurable without inference with the cell's activities. In this paper we consider two electrically coupled neurons. We present an approach to reconstruct the input current into one of the neurons using concepts from control theory. More precisely, we employ an unknown input observer to reconstruct the internal states of the neuron's model. A crucial part of this approach is the transformation of the nonlinear model into an appropriate normal form. Additional filtering yields the desired excitation current.