The diagnosis of invasive aspergillosis (IA) has always been a challenge in immunocompromised patients. Until recently, classical mycological tests were the only diagnostic tools available in this highly lethal condition. However, fungal culture and antibody detection are known for having a low diagnostic sensitivity. Histopathological examination is difficult, because biopsies in such immunocompromised hosts are risky and commonly not obtained, due to low platelet counts. Galactomannan (GM) testing has completely modified this scenario, allowing for an early, sensitive (78 %) and specific (81 %) diagnosis of IA, particularly in the neutropenic population [1]. In conjunction with computed tomography of the thorax, GM has become the backbone of preemptive antifungal therapy. Nowadays, most patients with IA enroll clinical trials or receive antifungal therapy based on GM testing. Nevertheless, there are a number of caveats that sometimes cloud the interpretation of GM results, including a high rate of false-positive results, cross-reaction with fungi other than the Aspergilli, and the reduced sensitivity of serum testing in nonneutropenic patients [2]. Therefore, the scientific community has been at a constant search for alternative non-culture-based and noninvasive diagnostic tests. Polymerase chain reaction (PCR) tests seem to fit quite well these requirements.The diagnostic performance of PCR in IA has been evaluated in several studies. A previous meta-analysis included 1618 at-risk patients from 16 studies to reveal that PCR had a sensitivity of 88 % in the diagnosis of IA [3]. Specificity was 75 % (increased to 87 % if two positive results are obtained). Most studies included in the meta-analysis used whole blood, and sample volume and DNA extraction protocols varied considerably, compromising reproducibility. More recently, a meta-analysis published at the Cochrane Library revealed that the mean sensitivity and specificity of Aspergillus PCR were 80 and 78 % for single positive test results, and 58 % and 96 %, respectively, for two consecutive positive results [4]-a performance very similar to what is known for GM. The elevated negative predictive value of Aspergillus PCR allows for IA to be virtually excluded in the presence of a negative test result. The performance of GM and PCR has been compared in a recent publication. The study showed that the highest diagnostic odds ratio (DOR) was obtained when GM was combined with PCR