Transcription factors are proteins which bind to deoxyribonucleic acid (DNA) and regulate gene expression, i.e. synthesis of a gene product. Understanding the mechanisms that regulate the expression of genes has been a major challenge in molecular biology. The identification of binding sites to DNA for transcription factors is a very important task in this challenge. These binding sites are known as motifs, and they are short DNA segments. The use of computational methods for prediction of these motifs is very effective. Several computational methods have been developed over the past few years for motif finding in DNA sequences. These techniques have also been extended to find motifs in proteins. For proteins, a motif is an amino-acid recurring pattern. But one of the major drawbacks of currently available methods is that the execution time is very high for large DNA or protein data sets. A hardware implementation of these computational methods would be a good alternative to reduce execution time.In this thesis, we have implemented the Expectation Maximization (EM) algorithm in hardware. EM is one of the widely used algorithms for motif finding. The entire hardware design has been realized using Verilog HDL modules. These modules can also be synthesized to generate gate-level netlists and be ported onto a field programmable gate array (FPGA). The functionality and performance of the design is tested on multiple data sets of varying sequence lengths. The performance of the design has also been compared to one of the popular software approaches, MEME. The tests show that the hardware approach can achieve speedups by a factor of 100 or greater.