Whole-genome DNA amplification (WGA) is a promising method that generates large amounts of DNA from samples of limited quantity. We investigated the accuracy of a multiplex PCR approach to WGA over STIR loci. The amplification bias within a locus and over all analyzed loci was investigated in relation to the amount of template in the WGA reaction, the specific STIR locus, and allele length. We observed reproducible error-free STIR profiles with 10 ng down to 1 ng of DNA template. The amplification deviation at a locus and between loci was within the intra-method reproducibility. WGA is the method of choice for amplifying nanograrn amounts of genomic DNA for different applications. We detected unbalanced STIR amplifications at one locus and between loci, allelic dropouts, and additional alleles after WGA of low-copy-number DNA. We found that the high number of drop-outs and drop-ins could be eradicated using pooled DNA from separate WGA reactions even with as little as 100 pg of starting template. Nevertheless, the quality of the results was still not sufficient for use in routine chimerism analysis of limited specific cell populations after allogeneic stem cell transplantation.