Effects of PK 11195 and flumazenil on cardiac responses to diazepam, clonazepam and zolpidem were compared. Coronary flow rate was increased at relatively low doses of diazepam and decreased at higher doses. Clonazepam induced a dose-dependent increase, and zolpidem a decrease of coronary flow rate. PK 11195 reduced the diazepam-induced increase of coronary flow rate, and flumazenil was ineffective. Neither antagonist evoked substantial changes in the decrease of coronary flow rate. PK 11195, and less so flumazenil, antagonized the clonazepam-induced increase. PK 11195 and flumazenil only in their highest doses suppressed and respectively potentiated the zolpidem-induced decrease. Inotropy showed a biphasic response in the presence of diazepam, i.e. an initial transient decrease, followed by a dose-dependent increase in two steps. Clonazepam induced a similar response. Zolpidem increased the inotropy. The negative inotropic response induced by diazepam did not change significantly in the presence of PK 11195 or flumazenil. The positive inotropic response was suppressed by PK 11195, and less so by flumazenil. The negative response to clonazepam was antagonized by both PK 11195 and flumazenil; the positive response was not significantly changed. In the presence of lower doses of PK 11195, the zolpidem-induced response was potentiated, whereas higher doses produced reversal; flumazenil potentiated the response. In conclusion, the results support earlier suggestions, involving receptor mechanisms with cardiac effects of benzodiazepines. Both agonists and antagonists (inter)act in a different manner, suggesting that rather ambiguous receptor mechanisms are involved in benzodiazepine effects in the heart.