In some arthropods there is a discrepancy in the number of dorsal tergites compared to the number of ventral sternites and leg pairs. The posterior tergites of the Diplopoda (millipedes) each cover two sternites and two pairs of legs. This segment arrangement is called diplosegmentation. The molecular nature of diplosegmentation is still unknown. There are even conflicting theories on the way the tergites and sternites/leg pairs should be correlated to each other. The different theories are based either on embryological analyses or on studies of the adult morphology and turned out to be not compatible with each other. We have previously used the expression patterns of segmentation genes in the pill millipede Glomeris marginata (Myriapoda: Diplopoda) to study millipede segmentation. Here we review the existing models on the alignment of tergites and leg pairs in millipedes with special emphasis on the implications the gene expression data have on the debate of tergite and leg pair assignment in millipedes. The remarkable outcome of the gene expression analysis was that (1) there is no coupling of dorsal and ventral segmentation and, importantly, that (2) the boundaries delimiting the tergites do neither correlate to the embryonic boundaries of the dorsal embryonic segments nor to the boundaries of the ventral embryonic segments. Using these new insights, we critically reinvestigated the correlation of tergites, sternites, and leg pairs in millipedes. Our model, which takes into account that the tergite boundaries are different from the dorsal embryonic segment boundaries, provides a solution of the problem of tergite to sternite/leg pair correlation in basal milipedes with non-fused exoskeletal elements and also has implications for derived species with exoskeletal rings. Moreover, lack of coupling of dorsal and ventral segmentation may also explain the discrepancy in numbers of dorsal tergites and ventral leg pairs seen in some other arthropods.