The present study examined the pharmacological actions of four different plant-derived essential oils (rose, ylang-ylang, camomile, orange) in two types of conflict tests using ICR mice. In the Vogel conflict test, in which any drinking behavior of the mice was punished by an electric shock, the benzodiazepine agonist, diazepam (DZ), increased the number of electric shocks the mice received. This number increased after administration of rose oil. In contrast, ylang-ylang, camomile, and orange oil did not produce such an effect in this test. In the Geller conflict test where lever-pressing of mice was reinforced by food pellets and then punished by electric shock, response (lever-pressing) rate during the alarm period was increased as well by the positive control drug, DZ. Similarly, the response rate during the alarm period increased after administration of rose oil. Here as well, ylang-ylang, camomile, and orange oils did not produce an anticonflict effect. In the Vogel conflict test, the anticonflict effect of DZ was reversed by the benzodiazepine antagonist, flumazenil (Ro15-1788) (FL). However, the effect of rose oil in this test was not antagonized by FL. The present study showed that rose oil possesses anticonflict effects, and that the effects are not mediated by the benzodiazepine binding site of the GABA(A) receptor complex. Such pharmacological actions may at least partially account for human behavioral effects attributed to essential oils.