A phylogenetic tree based on ITS1, Adh1 and ndhF grouped the species of the genus Sorghum into one distinct monophyletic group, but including two sister lineages, one with x=5, the other with x=10 as basic chromosome numbers. The goal of this study was to elucidate major patterns in Sorghum genome evolution, particularly n=5 vs. n=10 genomes. A very recent molecular cytogenetic study in our laboratory revealed striking structural karyotypic rearrangements between S. bicolor (x=10) and an x=5 Sorghum species, S. angustum; so an immediate objective here was to determine if identical or similar rearrangements exist in other wild Sorghum species. Our approach was [1] to extend similar methods to additional species, i.e., fluorescent in situ hybridization (FISH) analyses of sorghum genomic bacterial artificial chromosome clones and multi-BAC cocktail probes to mitotic chromosomes of S. angustum, S. versicolor, S. brachypodum and S. intrans; and [2] to augment the BAC-FISH findings by comparing telomeric and ribosomal DNA FISH signal distributions to x=5 and x=10 Sorghum species. Signals from in situ hybridizations of BAC-based probes were insufficiently robust and insufficiently localized to delineate FISH signal patterns akin to those discovered previously in S. angustum. Southern blots of the same BACs to restricted DNA of these species revealed relatively moderate affinity to smeared DNA, suggesting homology to non-tandemized sequences. FISH of the A-type TRS (Arabidopsis-like telomeric repeat sequence) revealed its presence is limited to terminal chromosomal regions of the Sorghum species tested, except S. brachypodum, which displayed intercalary signal on one chromosome and no detachable signal at its termini region. The hybridization of 45S and 5S rDNA revealed that the respective sites of tandemized clusters differ among species in terms of size, number and location, except S. angustum versus S. versicolor. Well localized BAC-FISH signals normally occur when signals from low-copy sequences discernibly exceed background signal, including those from hybridization of dispersed repetitive elements. The low level of signal intensity from BAC low-copy sequences relative to the background signal "noise" seems most likely due to low homology and(or) technical constraints. Extensive dispersal of low-copy sequences that are syntenic in S. bicolor seems unlikely, but possible. In conclusion, the result was a lack of clear experimental success with BAC-FISH and an inability to effectively screen for S. angustum-like rearrangements using BAC-FISH. The telomeric and rDNA FISH indicated that the x=5 genomes vary extensively. One can surmise that although the arrangements seen in S. angustum might extend to S. versicolor, they certainly do not extend to S. versicolor, they certainly do not extend to S. intrans or S. brachypodum. It is clear that S. brachypodum has telomeric repeats that are either very short or rely on some sequence other than the A-type TRS.