Linkage of bipolar disease to several markers mapping to chromosome 4p has been reported in an extended family multiply affected with bipolar affective disorder and no linkage was found at other locations with 106 microsatellite markers, of which 58 were dinucleotide and 48 tetranucleotide repeats [Blackwood et al. (1996), Nature Genetics, 12, 427-430]. Collecting these data provided the opportunity to assess the usefulness and accuracy of the automated linkage preprocessor (ALP) programme in a linkage study and to make a detailed comparison of di- and tetranucleotides with this semi-automated system. Genotypes were acquired using the automated linkage preprocessor (ALP) without any manual intervention at any stage of the procedure and results of analyses of these data were compared with results based on genotypes checked by visual inspection of the data. The ALP program was found to be timesaving and reliable and yielded similar results to non-automated reading using both di- and tetranucleotide repeat microsatellite markers. Tetranucleotides had fewer errors due to multiple genotypes and a lower incidence of stutter peaks making them more informative than dinucleotides in this linkage study.