This dissertation presents an investigation of micromachined three-dimensional microelectrode arrays (3-D MEAs) targeted toward in-vitro and in-vivo biomedical applications. Current 3-D MEAs are predominantly silicon-based, fabricated in a planar fashion, and are assembled to achieve a true 3-D form: a technique that cannot be extended to micro-manufacturing. The integrated 3-D MEAs developed in this work are polymer-based and thus offer potential for large-scale, high volume manufacturing. Two different techniques are developed for microfabrication of these MEAs - laser micromachining of a conformally deposited polymer on a non-planar surface to create 3-D molds for metal electrodeposition; and metal transfer micromolding, where functional metal layers are transferred from one polymer to another during the process of micromolding thus eliminating the need for complex and non-repeatable 3-D lithography processes. In-vitro and in-vivo 3-D MEAs are microfabricated using these techniques and are packaged utilizing Printed Circuit Boards (PCB) or other low-cost manufacturing techniques. To demonstrate in-vitro applications, growth of 3-D co-cultures of neurons/astrocytes and tissue-slice electrophysiology with brain tissue of rat pups were implemented. To demonstrate in-vivo application, measurements of nerve conduction were implemented. Microelectrode impedance models, noise models and various process models were evaluated. The results confirmed biocompatibility of the polymers involved, acceptable impedance range and noise of the microelectrodes, and potential to improve upon an archaic clinical diagnostic application utilizing these 3-D MEAs.