1. The occurrence of genetic adaptation to heavy metals in natural populations of terrestrial invertebrates is evaluated from literature data. Five criteria for adaptation evidence are applied, with concepts from ecotoxicology, ecology, life-history theory and quantitative genetics. 2. There is strong evidence for the occurrence of adaptation in natural populations of the isopod Porcellio scaber (Isopoda), the springtails Isotoma notabilis, Onychiurus armatus and Orchesella cincta (Collembola), the blowfly Lucilia cuprina and the fruit fly Drosophila melanogaster (Diptera). Adaptation to metal-containing pesticides has been demonstrated in ticks (Acarina). Population divergence indicates acclimation or adaptation in many other species. 3. Metal adaptation has been achieved within a few generations under laboratory conditions in some species; adapted populations occur at field sites that have been polluted for decades, or longer. 4. Genetic variation for tolerance and life-history characteristics, allowing for adaptation, was quantified in a reference population of Orchesella cincta. Tolerance and life-history patterns in exposed field populations matched predictions from genetic variation. 5. Adaptation involves modification and intensification of existing physiological mechanisms for metal assimilation, excretion, immobilization or compartmentalized storage. There are indications of inter-population divergence in metal-binding proteins in a snail. In the fruit fly Drosophila melanogaster metal adaptation is achieved by duplication of the metallothionein gene. 6. An altered life-history is often part of the complex adaptation syndrome. Metal-adapted invertebrates have a shorter life-cycle and a higher reproductive effort. 7. Possible consequences of adaptation, consisting of costs of tolerance determined by genetic correlations, and probably of reduced genetic variation for tolerance and other features, are discussed. Reduced genetic variation is suggested by results for the springtail Orchesella cincta. 8. The distinction between ''costs of tolerance'' on the one hand and linkage disequilibrium or direct selection for altered life-history patterns on the other hand is discussed. 9. Species with high sensitivity (i.e. a low NOEC), that do not have populations maintaining sufficient genetic variation to evolve tolerance or modified life-history characteristics, or that have costly tolerance mechanisms, or both, are most at risk for extinction at sites with increasing metal pollution. 10. Metal adaptation in terrestrial invertebrates appears to be of degree rather than of kind: indications for a specific metal-fauna, equivalent to metal-vegetation, are lacking.
