Multiple sclerosis (MS) is a demyelinating disease of the central nervous system with a complex genetic background. In order to identify loci associated with the disease, we had performed a genome screen initially using 6000 microsatellite markers in pooled DNA samples of 198 MS patients and 198 controls. Here, we report on the detailed reanalysis of this set of data. Distinctive features of microsatellites genotyped in pooled DNA causing false-positive association or masking existing association were met by improved evaluation and refined correction factors in the statistical analyses. In order to assess potential errors introduced by DNA pooling and genotyping, we resurveyed the experiment in a subset of microsatellite markers using de novo-composed DNA pools. True MS associations of markers were verified via genotyping all individual DNA samples comprised in the pools. Microsatellites share characteristically superb information content but they do not lend themselves to automation in very large scale formats. Especially after DNA pooling many artifacts of individual marker systems require special attention and treatment. Therefore, in the near future comprehensive whole-genome screens may rather be performed by typing single nucleotide polymorphisms on chip-based platforms.