Nonnative species are being introduced to lakes and rivers at increasing rates worldwide. The impacts of the vast majority of these invasions have not been studied. Most of these species are thought to have only minor ecological effects, whereas others are known to have substantially altered water quality, contaminant cycling, food webs, and native biodiversity. Managers lack risk assessment methods to prioritize invasion threats, because there exist very few general models, or even ‘rules of thumb’, to predict the impacts of aquatic invasions. A major challenge to prediction is the variation generated by the influence of site-specific physical and biological factors that affect the invader’s local abundance and performance. In particular, interactions with other introduced species can produce synergistic effects that are extremely difficult to predict. Nevertheless, analyses of data from invasions worldwide have revealed patterns of impact that may prove to be useful to risk assessment. Invaders that have strong impacts typically have the following characteristics: (1) they have high reproductive rates; (2) they are introduced into regions where no ecologically similar organisms exist; (3) they are either predators or suspension feeders, with high rates of resource consumption; and (4) they have a history of impacts in other regions. Other caveats have emerged: the ‘tens rule’ (that only about 10% of established species become harmful ecologically or economically) may severely underestimate the true proportion of species that threaten fisheries, and there appears to be no relationship between the impact of an introduced species and its ability to spread.