The analysis of gene expression data in DNA chips is an important tool used in genomic research whose main objectives range from the study of the functionality of specific genes and their participation in biological process to the reconstruction of diseases's conditions and their subsequent prognosis. Gene expression data are arranged in matrices where each gene corresponds to one row and every column represents one specific experimental condition. The biclustering techniques have the purpose of finding subsets of genes that show similar activity patterns under a subset of conditions. Our approach consists of a biclustering algorithm based on local nearness. The algorithm searches for biclusters in a greedy fashion, starting with two–genes biclusters and including as much as possible depending on a distance threshold which guarantees the similarity of gene behaviors.