Background: Little is known about the development of cross-reactive antibodies following natural exposure to pathogens. Such knowledge is critical in the development of new universal influenza vaccines.Methods: To study the possibility of the presence of cross-reactive antibodies to influenza viruses which underwent a major antigenic drift between the years 1999 and 2007 sera from samples of 80 children and 400 adults were selected at random from the Israeli national serum bank. The sera was obtained in 2002 and in 2007, two time points that followed a major drift in the infectious H3N2 influenza virus strain (A/Panama/2007/99 to A/Wisconsin/67/2005).Results: In the summer of 2002, 13% of the children had Hemagglutination Inhibition (HI) antibody titers of at least 40 and these antibodies recognized both A/Panama/2007/99 and A/Wisconsin/67/2005, where the latter strain only began to circulate in Israel in 2006. In 2007, 29% of the children had HI antibody titers of at least 40 directed against both A/Wisconsin/67/2005 and A/Panama/2007/99, even though they had never been exposed to the latter virus. Anti-A/Panama/2007/99 antibodies were detected in 58% and 68% of the 2002 and 2007 adult samples, respectively, while 8% and 39% had antibodies against A/Wisconsin/67/2005, respectively.Conclusions: The presence of naturally occurring cross-reactive influenza virus antibodies in a significant percentage of children has important implications for the development of a universal influenza vaccine.