Published indirect evidence indicating that the toxicity of solutions of complex metal cyanides to fish is a function of the concentration of molecular (not ionized) HCN is reviewed. The striking influence of pH on the toxicity of the nickelocyanide complex could be explained only by assuming that HCN is the toxic factor. Recent re-evaluation of the dissociation constant of the tetracyanonickelate (II) ion, which showed previously accepted values to be grossly inaccurate and the true value to be in close agreement with an estimate based on toxicological data and theory, has confirmed the above interpretation of toxicity test results.