Short-term exposure of plants to heavy metals is often used for risk assessment of metal-enriched soils (OECD guideline 208) without considering the reliability of the assessment for long-term exposure, i.e. for the completion of a plant's life-cycle. In the present study with 15 orogenic soils three phases of the life-cycle of a Zn-Cd-resistant ecotype of Silene vulgaris were studied to improve risk assessment of metal-enriched soils. The first phase, i.e. emergence of seedlings was not related to the water-soluble or total metal concentration of the soils. Seedling mortality was low as long as the water-soluble metal concentration did not surpass 0.15 micromol Zn and 0.04 micromol Cu g(-1) dry soil. Curtailment of the life-cycle prior to flowering, i.e. the vegetative growth as second phase, occurred on those soils where roots and shoots were heavily enriched by Zn already in the seedling phase. In the third phase, i.e. the generative phase, time to flowering and yield differences between orogenic soils were substantial, but soil metal concentrations could not be directly related to timing of reproduction or biomass. Ranking of data showed a high inconsistency of the responses to metal exposure during the first phases of the life-cycle. It is concluded that total plant mass and seed mass are the only realistic endpoints of life-cycle bioassays in risk assessment as long as ranks are inconsistent between two successive early phases of the life-cycle.