ABSTRACT We have discovered an easy way to cut through the mitotic spindle at any desired place. Spindles of demembranated cricket or grasshopper spermatocytes were severed with a microneedle between the chromosomes and one pole, and the cut-off polar piece was swept away. Spindle structure and microtubule dynamics in cut spindles were studied by anti-tubulin immunostaining and electron microscopy. The cut is clean: all microtubules are severed and only a few extend beyond the others. This provides the basis for a clear test of whether traction fibers pull chromosomes to the pole in anaphase, because the putative traction fiber is cleanly severed. Cutting creates new plus ends on microtubules in the cut-off polar piece and new minus ends on microtubules in the main spindle body. The microtubules with new plus ends are unstable, as expected from the dynamic instability of microtubules. However, the microtubules with new minus ends are as stable as uncut microtubules in the same spindle. Our mechanical method of cutting microtubules very likely creates native, reactive ends, and therefore the surprising stability of new minus ends is genuinely interesting, not an artifact of cutting.