Twenty-one Citrus tristeza virus (CTV) isolates from India were characterized, using genotype-specific multiple molecular markers (MMM) from the 54'-terminal region and two other overlapping primer pairs (CN487/489 and CN488/491) from ORF1a (697-1484 nucleotides (nt)). The 5'-terminal genotype-specific primer pairs amplified about 500 bases from the 5'-end of the CTV genomic RNA (gRNA). With the three different MMM, the VT genotype-specific primers amplified 19 Indian CTV isolates. The T30-specific primers amplified five isolates, and the T36 primer amplified only one isolate T36. All isolates were amplified with CN488/491 primers; however, only 20 isolates were amplified with CN487/489 pair. A phylogenetic tree, derived from the sequences of the different MMM primer-amplified products, placed all the isolates into four distinct genogroups. Three of these four groups were typified by the reference isolates T30, T36, and VT. The fourth group, represented by the isolate BAN-2, was considered as a new genogroup. A phylogenetic tree based on sequences of the CN487/491 amplified products and other published sequences placed all of the isolates in eight genogroups. Phylogenetic correlation over the three different regions sequences of these CTV isolates showed more sequence variability between 1082 and 1484nt than between 1 and 500 or 697-1105 nt of the CTV gRNA. Based on three different 5' regions sequences and phylogenetic analysis, it is hypothesized that isolates BAN-1, BAN-2, and B165 are three naturally occurring variants that add to the complexity of the CTV populations in India.