The aim of this paper is to find the location of a signal source using time difference of arrival. The paper first defines time difference of arrival and how it can give us a unique solution. It proceeds to present two estimation methods – Friedlander’s Time Difference of Arrival Algorithm and Newton’s Method. The first method provides an initial solution, while the second method attempts to improve on the initial solution. After deriving the algorithm, we attempted a simulation of time difference of arrival using an audio file, an arbitrary number of sensors whose locations are known, and a signal source whose location is also known. An attempt at simulating the effects of different number of sensors, the distance between sensors and source location, the sample rate, the signal to noise ratio, and, finally, different sensor placements on our accuracy has also been conducted. In order to confirm the accuracy, the algorithm has been tested using real microphones. The analysis of time difference of arrival has produced the following results. The results obtained through a simulation of time difference of arrival can be precise, even when the sensors are close to each other. Cross correlation can handle signals with a negative SNR. Whereas increasing the sample rate can drastically improve our results, adding more sensors does not necessarily improve the accuracy. For the accuracy to be improved, the sensors should be placed in optimal locations.