As micronuclei (MN) derive from chromosomal fragments and whole chromosomes lagging behind in anaphase, the MN assay can be used to show both clastogenic and aneugenic effects. The distinction between these phenomena is important, since the exposure studied often induces only one type of MN. This particularly concerns the use of MN as a biomarker of genotoxic exposure and effects, where differences in MN frequencies between exposed subjects and referents are expected to be small. A specific analysis of the induced type of MN may considerably improve the sensitivity of detecting the exposure effect. MN harbouring chromosomes can be distinguished from those harbouring acentric fragments by the presence of a centromere. The proportion of centromere-positive MN in human lymphocytes increases with age, which primarily reflects an age-dependent micronucleation of the X and Y chromosomes. The X chromosome especially tends to lag behind in female lymphocyte anaphase, being micronucleated more efficiently than autosomes. There is some evidence for an enhanced prevalence of fragments from chromosome 9 in spontaneous human lymphocyte MN and from chromosomes 1, 9 or 16 in MN induced in vitro by some clastogens; the breakage appears to occur in the heterochromatic block of these chromosomes. Although there are indications that centromere identification can improve the detection of clastogenic effects in humans in vivo, smokers have not shown an increase in centromere-negative MN in their cultured lymphocytes, although smoking is known to produce chromosomal aberrations. This may suggest that fragment-containing MN and chromosomal aberrations cover partly different phenomena. Understanding the mechanistic origin and contents of MN is essential for the proper use of this cytogenetic end-point in biomarker studies, genotoxicity testing and risk assessment.