The hemagglutinins of A Shangdong/9/93 (S93 [H3N2]) and A/Beijing/32/92 (B32 [H3N2]) were identified by serologic responses with animal and human sera to be markedly different from previous H3N2 strains, and therefore candidates for new vaccines. Reassortant influenza viruses with increased ability to replicate in eggs were produces using A/Puerto Rico/8/34 (PR8 [H1N1]) combined with S93 or B32 to provide surface glycoproteins derived from the hemagglutinin (HA) and neuraminidase (NA) genes of the wild type strains. In order to improve certainty of identification of the origin of individual genes in the reassortants, a modified polymerase chain reaction method was developed so that all eight viral gene segments could be amplified simultaneously with one set of oligonucleotide primers for both reverse transcription of the viral RNA and subsequent amplification of the complementary DNA. The relative efficiency of amplification of larger gene segments was markedly improved by adjustment of dimenthylsulfoxide added to the reaction mixture, so that nearly equimolar concentrations of each gene segment could be recovered. Although HA and NA genes are sufficiently different for positive identification of strain of origin by mobility in agarose, the other six genes are less variable among most influenza A strains. Therefore, restriction mapping of individual gene segments was developed for key genes. For example, restriction enzymes with specificity for a single site in gene segment 7 ([GS7]) one of the genes critical to replication in eggs) were identified for PR8, for modern H1N1 strains and for current H3N2 strains to permit positive identification of the strain of origin. As a supplement to the restriction site analyses, the sequence of the variable 5~ end of GS7 was determined. In comparison to the sequences reported by two other labs for GS7 from PR8, the strain of PR8 used here for reassorting is intermediate between the published sequences. Similar sequence analysis for A/Beijing/353/89 and B92 (which represent recent change in the H3N2 HA) in comparison to published information for H3N2 strains isolated during early and late 1970~s indicated that GS7 in H3N2 strains has well been conserved over 25 years. The rapidity, sensitivity, and relative simplicity of the methods suits them to preparing influenza reassortants for vaccines and to identifying potential mixed populations or contamination of influenza virus strains. The strategy of restriction mapping and sequencing variable genomic regions is being extended to other genes.