The mechanism of nonrandom chromosome segregation and elimination in male first meiotic division of sciarid flies is one of the enigmas of classical cytogenetics. Interpretations of the available data regarding this unusual chromosome transport are contradictory [reviewed by Gerbi, 1986: “Germline—Soma Differentiations. Results and Problems of Cell Differentiation,” Vol. 13:71–104]. In the present study of spermatogenesis in Trichosia pubescens the process is reinvestigated by indirect anti‐tubulin immunofluorescence labeling of spindle microtubules and DAPI chromosome staining, partly in combination with ultrathin sectioning and electron microscopy. While the maternal homologues and the sex‐limited (L) chromosomes seem to be transported very quickly toward the pole of the monopolar spindle, where they stay throughout first meiotic division, paternal homologues remain some distance from the pole, stick together, and, as an undifferentiated cluster of chromatin, withdraw from the pole in the course of division. Finally the paternal chromatin becomes eliminated in a cytoplasmic bud which is cast off. The different behavior of maternal homologues (and L chromosomes) and paternal homologues may be caused by differences in kinetochore structure and function. In contrast to maternal and L chromosomes, the paternal homologues do not display structurally defined kinetochore‐like regions and seem to be unable to orient poleward. During the process of elimination, a prominent bundle of pole‐oriented microtubules is associated with the paternal chromatin cluster. It is suggested that transport away from the pole is brought about by these microtubules. Thus, meiotic chromosome elimination in sciarids may be related to mechanisms involving “polar ejection forces” [Rieder et al., 1986: J. Cell Biol. 103:581–591]. Cell Motil. Cytoskeleton 36:84–94, 1997. © 1997 Wiley‐Liss, Inc.