A challenge in ecotoxicology is the translation of individual-level effects of exposure, as measured in for instance laboratory tests, into population-level effects that have more ecological relevance. It is not necessarily possible to predict effects of toxicant exposure on the population level directly from observations on individual-level endpoints. Density-dependent processes within natural populations is a major complicating factor. In this study, laboratory populations of zebrafish (Danio rerio) were exposed to a naturally occurring mixture of persistent organic pollutants (POPs) at three levels of population density. It was found that toxicant exposure decreased somatic growth in juveniles and adults, and increased time to sexual maturation. Increasing population density had the same effects. For fish populations, individual growth can be an important regulatory mechanism. Size-structured matrix population models were therefore developed in order to synthesise the toxicant effects on vital rates into population-level endpoints. Subsequently, two sets of simulations were run, one with a density-independent model and one with a model where larval survival, juvenile and adult growth and fecundity were modeled as functions of population density. The outcome of the simulations with the density-independent model indicated that exposure to POPs may result in reduced population growth rate and reduced population abundance. When density dependence was taken into account in the model, the simulation results indicated that under certain regimes of density dependence, toxicant exposure has the potential to relieve effects of density stress on individual-level vital rates. Measures of population growth rate and/or abundance may be difficult to obtain in an experimental setting, and thus population models are important tools in ecological risk assessment of toxicants. Further analysis of the models developed in this thesis should include sensitivity and elasticity analysis. These types of analyses may provide valuable information about the magnitudes of the population-level impacts of given toxicant-induced reductions in the different vital rates.