Despite PCR per se being a powerful and sensitive technique, regarding the detection of fungi in patients' blood, no consensus for a standardised PCR protocol yet exists. To complement other ongoing or accomplished studies which tackle this problem, the German Reference Center for Systemic Mycoses conducted an interlaboratory comparison starting with blood samples spiked with fungal cell elements. Altogether, six laboratories using in-house PCR-protocols from Germany and Austria participated in the trial. Blood samples were spiked with vital cells of Candida albicans or Aspergillus fumigatus. Candida was used in the yeast form, whereas Aspergillus cells were either spiked as conidia or as very young germlings, also known as smoo cells. Spiked blood samples contained between 10 and 10 000 cells ml(-1). Depending on the techniques used for fungal cell disruption and DNA-amplification, detection quality was variable between laboratories, but also differed within single laboratories in different trials particularly for samples spiked with less than 100 cells ml(-1). Altogether, at least regarding the detection of A. fumigatus, two of six laboratories showed constant reliable test results also with low fungal cell number spiked samples. Protocols used by these labs do not differ substantially from others. However, as particularities, one protocol included a conventional phenol chloroform extraction during the DNA preparation process and the other included a real time PCR-protocol based on FRET probes. Other laboratory comparisons on the basis of clinical samples should follow to further evaluate the procedures. The difficulties and problems of such trials in general are discussed.