Influenza A virus of the Orthomyxoviridae family is a contagious respiratory pathogen that continues to evolve and burden in the human public health. It is able to spread efficiently from human to human and have the potential to cause pandemics with significant morbidity and mortality. It has been estimated that every year about 500 million people are infected with this virus, causing about approximately 0.25 to 0.5 million people deaths worldwide. Influenza A viruses are classified into different subtypes by antigenicity based on their hemagglutinin (HA) and neuraminidase (NA) proteins. The sudden emergence of influenza A virus subtypes and access for epidemiological analysis of this subtypes demanded a rapid development of specific diagnostic tools. Also, rapid identification of the subtypes can help to determine the antiviral treatment, because the different subtypes have a different antiviral drug resistance patterns. In this study, our aim is to detect influenza A virus subtypes by using real-time nucleic acid sequence based amplification (NASBA) which has high sensitivity and specificity through molecular beacon. Real-time NASBA is a method that able to shorten the time compare to other molecular diagnostic tools and is performed by isothermal condition. We selected major pandemic influenza A virus subtypes, H3N2 and H5N1. Three influenza A virus gene fragments such as HA, NA and matrix protein (M) gene were targeted. M gene is distinguished influenza A virus from other influenza virus. We designed specific primers and molecular beacons for HA, NA and M gene, respectively. In brief, the results showed that the specificity of the real-time NASBA was higher than reverse transcription polymerase chain reaction (RT-PCR). In addition, time to positivity (TTP) of this method was shorter than real-time PCR. This study suggests that the rapid detection of neo-appearance pandemic influenza A virus using real-time NASBA has the potential to determine the subtypes.