OBJECTIVE To describe, validate and apply a computer simulation of corpuscle sectioning to determine which assumptions are compatible with the possibility of correction. STUDY DESIGN A computer program was written in PASCAL and run on a personal computer. RESULTS Correction is practical in the case of spherical nuclei with homogeneous intranuclear DNA and a constant internuclear DNA concentration, even if section thickness is uneven or inaccurate. Correction is not possible in the case of marked nuclear ellipticity: mild cases are correctable, although subpopulations in a mixed-ploidy sample would not be distinguished. Correction is possible in the case of spherical (but not ellipsoidal) nuclei with a variable internuclear DNA concentration, and subpopulations in a mixed-ploidy sample would be distinguishable if sufficiently different. If the DNA is markedly concentrated in one part of the nucleus, overcorrection results. CONCLUSION The results clarify the factors that limit correction of image cytometric measurements of ploidy in tissue sections. A new method of correcting ploidy measurements in tissue sections is usable.