The aim of these experiments was to evaluate how thresholds for phytotoxic substances obtained in seedling bioassays relate to yield losses or changes in yield components of mature barley crops after a short-term exposure to p-hydroxybenzoic acid. Under laboratory conditions a treatment with 1.81 mM p-hydroxybenzoic acid significantly reduced the radicle length of barley, whereas coleoptile elongation was less sensitive. The inhibition of the radicle length and coleoptile elongation was greater if the pH of the test solution was not buffered at pH 5.5. In a glasshouse trial the effect of p-hydroxybenzoic acid on the radicle and coleoptile elongation of spring barley was compared with the yield response after a three day exposure either during germination or at the double ridge stage of apex development. Applications of 0.72mM, 1.44mM and 3.62mM p-hydroxybenzoic acid averaged over the treatments during germination or at the double ridge stage of development caused a yield reduction in the single ear weight of 5%, 13% and 19% in comparison with the control, respectively. The higher tiller categories in general showed a greater sensitivity towards an application of p-hydroxybenzoic acid and, therefore, could not compensate for the yield decrease of the main stem tiller. A single application of phydroxybenzoic acid either at germination or at the double ridge stage may cause yield losses, as reported from no-till systems or cereal monocultures. The data have implications for the interpretation of seedlings bioassays in allelopathic research and their applicability in estimating yield losses caused by phytotoxic substances.