Recent years have witnessed some convergence of opinion regarding the identity of the rhizotoxic species of aluminium, AlO4Al12(OH)(24)(H2O)(12)(7+) (Al-13) and Al3+ are almost certainly toxic, but no rhizotoxicity has been detected for Al-SO4 (AlSO4+ and Al(SO4)(2)(-)) or Al-F (e.g. AlF2+ and AlF2+). The status of Al-OH (e.g. AlOH2+ and Al(OH)(2)(+)) is uncertain because experimental results often appear to indicate Al-OH toxicity, In this article it is argued that this appearance of toxicity is the consequence of the relief of Al3+ toxicity by H+ (and vice versa), Furthermore, this view provides an explanation for phenomena, such as Al stimulation of growth, unexplained by the hypothesis that Al-OH is toxic, It is concluded, therefore, that Al-OH is not toxic at achievable activities, The previous failure to detect toxicity from Al-SO4 and Al-F also may have been a consequence of the low activities used in the experiments. New experiments again failed to reveal a toxicity for Al-SO4, but do indicate that AlF2+ and AlF2+ are toxic, Several alternative hypotheses for the apparent toxicity of Al-F were considered and rejected, Consequently, this report concludes that the following Al species are toxic to wheat roots in the following order: Al-13 > Al3+ > AlF2+ > AlF2+. The correlation between toxicity and charge is discussed in light of the fact that the binding strength of cations to plasma membrane surfaces generally increases with charge.