In this letter to the editor, Sher notes a recent report that suggested low levels of cerbrospinal fluid 5-hydroxyindolacetic acid predicted completed suicide (Samuelsson et al, 2006). This observation is consistent with multiple reports suggesting serotonergic abnormalities are involved in the neurobiology of suicidal behaviour. It is asserted that it is reasonable to hypothesize cerebrospinal fluid & homovanillic levels in combination with other biological & clinical markers of suicidal behaviour may help predict & prevent suicide in some patient populations. (14 refs.)