Selection of an ideal sample is a vital element in early detection of influenza infection. Rapid identification of infectious individuals or animals is crucial not only for avian influenza virus (AIV) surveillance programmes, but also for treatment and containment strategies. This study used a combination of quantitative real-time RT-PCR with an internal positive control and a cell-titration system to examine the presence of virus in different samples during active experimental AIV infection and its persistence in the infected carcasses. Oropharyngeal/cloacal swabs as well as feather pulp and blood samples were collected from 15-day-old chicks infected with H7N1 highly pathogenic AIV (HPAIV) and the kinetics of virus shedding during active infection were evaluated. Additionally, several samples (muscle, skin, brain, feather pulp and oropharyngeal and cloacal swabs) were examined to assess the persistence of virus in the HPAIV-infected carcasses. Based on the results, feather pulp was found to be the best sample to detect and isolate HPAIV from infected chicks from 24 h after inoculation onwards. Kinetic studies on the persistence of virus in infected carcasses revealed that tissues such as muscle could potentially transmit infectious virus for 3 days post-mortem (p.m.), whilst other tissues such as skin, feather pulp and brain retained their infectivity for as long as 5-6 days p.m. at environmental temperature (22-23 degrees C). These results strongly favour feather as a useful sample for HPAIV diagnosis in infected chickens as well as in carcasses.