The aim of this study was to assess the acute toxicity of polycyclic aromatic hydrocarbons using lux-marked bacterial biosensors. Standard solutions of phenanthrene, pyrene and benzo[a]pyrene were produced using 50 mM hydroxpropyl-beta-cyclodextrin solution which contained each respective polycyclic aromatic hydrocarbon at 6.25 times the aqueous solubility limit of the compound. The polycyclic aromatic hydrocarbon solutions were incubated with each of the biosensors for 280 min and the bioluminescence monitored every 20 min. Over the incubation time period, there was no significant decrease in bioluminescence in any of the biosensors tested with the exception of Rhizobium leguminosarum biovar trifolii TA1 luxAB. In this series of incubations, there was a dramatic increase in bioluminescence in the presence of phenanthrene (2.5 times) and benzo[a]pyrene (3 times) above that of the background control (biosensor without polycyclic aromatic hydrocarbon) after 20 min. Over the next 3 h, bioluminescence decreased to that of the control. An ATP assay was carried out on the biosensors to assess if uncoupling of the oxidative phosphorylation mechanisms in the respiratory chain of the cells had occurred. However, it was found that the polycyclic aromatic hydrocarbons had no effect on the organisms indicating that there was no uncoupling. Additionally, mineralisation studies using 14C-labelled polycyclic aromatic hydrocarbons showed that the biosensors could not mineralise the compounds. This study has shown that the three polycyclic aromatic hydrocarbons tested are not acutely toxic to the prokaryotic biosensors tested, although acute toxicity has been shown in other bioassays. These results question the rationale for using prokaryote biosensors to assess the toxicity of hydrophobic chemicals, such as polycyclic aromatic hydrocarbons.