We report the development of metal transfer micromolded (MTM) three-dimensional microelectrode arrays (3D MEAs) for a transcutaneous nerve tracking application. The measurements of electrode–skin–electrode impedance (ESEI), electromyography (EMG) and nerve conduction utilizing these minimally invasive 3D MEAs are demonstrated in this paper. The 3D MEAs used in these measurements consist of a metalized micro-tower array that can penetrate the outer layers of the skin in a painless fashion and are fabricated using MTM technology. Two techniques, an inclined UV lithography approach and a double-side exposure of thick negative tone resist, have been developed to fabricate the 3D MEA master structure. The MEAs themselves are fabricated from the master structure utilizing micromolding techniques. Metal patterns are transferred during the micromolding process, thereby ensuring reduced process steps compared to traditional silicon-based approaches. These 3D MEAs have been packaged utilizing biocompatible Kapton® substrates. ESEI measurements have been carried out on test human subjects with standard commercial wet electrodes as a reference. The 3D MEAs demonstrate an order of magnitude lower ESEI (normalized to area) compared to wet electrodes for an area that is 12.56 times smaller. This compares well with other demonstrated approaches in literature. For a nerve tracking demonstration, we have chosen EMG and nerve conduction measurements on test human subjects. The 3D MEAs show 100% improvement in signal power and SNR/√area as compared to standard electrodes. They also demonstrate larger amplitude signals and faster rise times during nerve conduction measurements. We believe that this microfabrication and packaging approach scales well to large-area, high-density arrays required for applications like nerve tracking. This development will increase the stimulation and recording fidelity of skin surface electrodes, while increasing their spatial resolution by an order of magnitude or more. Although biopotential electrode systems are not without their challenges, the non-invasive access to neural information, along with the potential for automation with associated electronic and software development, is precisely what makes this technology an excellent candidate for the next generation in diagnostic, therapeutic, and prosthetic devices.