Based on a schematic model of karyotype evolution, a new methodology f
or G-, R-, or Q-banding pattern analysis was investigated. Banding pat
tern analysis essentially depends on the unidirectional alteration and
the randomness of the exchange sites of the AM-inversion. In karyotyp
es that evolved by AM-inversion and Robertsonian rearrangement, two ma
tching patterns appear; (1) tandem and (2) complementary matching patt
erns. The former is characteristic of a single lineage sharing the sam
e AM-inversions, and the latter appears in different lineages sharing
different AM-inversions, by which it is theoretically possible to dete
ct the ancestral karyotype and to reconstruct the karyotype phylogeny
(cladogram). In contrast, the evolutionary pathway cannot always be pe
rceived if karyotypes evolve only by Robertsonian rearrangement. The t
andem matching pattern does not always mean tandem fusion, but can be
interpreted as 'tandem fission' by a combination of AM-inversion and c
entric fission. Tandem fusion and MM-inversion often cause entangled m
atching patterns, and thus they interfere with banding pattern analysi
s. Some methodological problems inherent in the conventional banding p
attern analysis are highlighted, and suggested that such problems can
be minimized by using the karyograph method. The methodology of bandin
g pattern analysis proposed in the present paper will be applicable fo
r matching the chromosome map of genetic markers among different speci
es.