Rats were exposed to increasing concentrations of morphine hydrochloride (up to 0.4 mg/ml) in 5% w/v sucrose solution as their sole source of drinking water. Physical dependence was established as determined by the precipitation of withdrawal behaviour following administration of 1 mg/kg IP naloxone hydrochloride on day 23. The choice between either a 5% w/v sucrose solution or a 5% w/v sucrose solution containing 0.4 mg/ml morphine hydrochloride 4 days following withdrawal resulted in rats being categorized into two groups based on their respective consumption of the morphine-containing solution. The amount of morphine solution voluntarily consumed by approximately half the rats were sufficiently high as to lead to a relapse into physical dependence to morphine. The high preference for morphine shown by these rats could not be attributed to the taste of the morphine solution. Naive rats or rats exposed to a 5% w/v sucrose solution for 23 days failed to consume significant quantities of the morphine-containing solution when provided with a choice. The administration of either an IM slow-release formulation of 70.5 mg/kg haloperidol decanoate (= 50 mg/kg haloperidol) or 10 micrograms/kg IP ondansetron hydrochloride daily did not alter morphine ingestion in the high morphine-preferring rats.