Efficiently aligns DNA sequencing reads with a reference genome. SMALT employs a hash index of short words (< 21 nucleotides long), sampled at equidistant steps along the genomic reference sequences. For each read, potentially matching segments in the reference are identified from seed matches in the index and subsequently aligned with the read using a banded Smith-Waterman algorithm. The best gapped alignments of each read is reported including a score for the reliability of the best mapping. The user can adjust the trade-off between sensitivity and speed by tuning the length and spacing of the hashed words. A mode for the detection of split (chimeric) reads is provided. Multi-threaded program execution is supported. Mapping with SMALT involves two steps: First, a hash index has to be generated for the genomic reference sequences. Then the sequencing reads are mapped onto the reference using the index.