The emergence of the pandemic influenza virus A H1N1 has made fast and accurate diagnosis essential. However, few well-validated diagnostic techniques exist. The real-time RT-PCR developed by the Centers for Disease Control and Prevention (CDC) is the recommended technique. Our objective was to compare the CDC real-time RT-PCR assay, shell vial (SV), and conventional cell culture [with Madine-Darby canine kidney (MDCK) cells and A549] for the detection of pandemic influenza A H1N1 in hospitalized patients. We performed a prospective study comparing the efficacy of 5 diagnostic techniques (RTPCR, SV in A549, SV in MDCK, conventional cell culture in A549, and conventional cell culture in MDCK) using nasopharyngeal swabs from patients ≥18 years of age hospitalized with clinical symptoms of influenza at our institution. Detection of the virus by conventional culture was considered the gold standard. An "extended gold standard" was also used to recalculate validity values. The sensitivities, specificities, positive predictive values, and negative predictive values (NPVs) for the detection of influenza A H1N1, determined using conventional culture as the gold standard, were, respectively, as follows: RT-PCR: 95.6, 82.3, 78.3, 96.5%; SVA549: 91.2, 99.01, 98.4, 94.4%; SV-MDCK: 82.3, 100, 100, 89.4%; tube-A549: 94.12, 100, 100, 96.2%; tube-MDCK: 86.7, 100, 100, 91.9%. Sensitivities and NPVs using an extended gold standard were as follows: RT-PCR: 96.5%, 96.6%; SV-A549: 73.3%, 78.5%; SV-MDCK: 65.1%, 73.7%; tube-A549: 74.4%, 79.2%; tube-MDCK: 68.6%, 75.7%. The average time to detect pandemic influenza A H1N1 by RT-PCR, SV culture, and conventional culture was, respectively, 4 h, 48 h, and 7 days. Real-time RT-PCR displayed high sensitivity and specificity for the detection of influenza A H1N1 in adult patients when compared with conventional techniques. In addition, the A549 cell line was not inferior to the MDCK line.