Successful management of invasive weeds will require active attempts to prevent new introductions, vigilant detection of nascent populations and persistent efforts to eradicate the worst invaders. To achieve these objectives, invasion ecology offers five groups of complementary approaches. (i) Stochastic approaches allow probabilistic predictions about potential invaders based on initial population size, residence time and number of introduction attempts. (ii) Empirical taxon-specific approaches are based on previously documented invasions of particular taxa. (iii) Evaluations of the biological characters of non-invasive taxa and successful invaders give rise either to general or to habitat-specific screening procedures. (iv) Evaluation of environmental compatibility helps to predict whether a particular plant taxon can invade specific habitats. (v) Experimental approaches attempt to tease apart intrinsic and extrinsic factors underlying invasion success. An emerging theory of plant invasiveness based on biological characters has resulted in several rather robust predictions which are presented in this paper.