Early drug development requires tests for compound-induced neurotoxic effects, i.e., to investigate possible alterations of neuronal activity as a result of the test compound. In vivo and in vitro animal models transpire not to be overly predictive of neurotoxic effects in humans and furthermore are in contradiction to the efforts of the European Community to reduce the number of animal experiments. Consequently, alternatives to these animal model-based assays are currently being investigated. Human induced pluripotent stem cell (hiPSC)-derived neurons offer several advantages, including being of human origin and offering the possibility of developing disease models from patient-derived cells. The development of electrophysiological assays based on microelectrode array systems (MEA) allows one to study alterations of neuronal activity in samples of varying complexity ranging from single cells to neuronal networks. As a non-invasive method it supports not only acute but also long-term experiments for extended time periods. Here we describe how to record neuronal activity from neurons and provide exemplarily insights into a validation study for a commercially available hiPSC-derived neuronal cell type.