Opiate dependence in laboratory animals is commonly induced by two methods: (1) subcutaneous (s.c.) insertion of morphine pellets, and (2) daily injections of increasing doses of opiates. While both of these methods reliably induce opiate dependence, they do not allow one to discontinue, and subsequently reestablish steady state opiate plasma levels with minimal invasive procedures. We developed an "ON-OFF" gating device for repeatedly and non-invasively turning ON or OFF opiate delivery by standard osmotic minipumps. The reliability of this "device" was tested utilizing naloxone (NAL)-precipitated somatic signs of withdrawal, and body mass index (BMI) as measures of withdrawal. Rats were implanted with osmotic minipumps equipped with the gating device, containing heroin (2.66 mg per day). Three days after surgery, somatic signs of withdrawal were precipitated every 48 h by NAL (0.3mg/kg), with minipumps gated ON or OFF. For BMI, spontaneous withdrawal was repeatedly (three times) induced by turning OFF and ON the gating devices every 48 h. Body weights were measured every 4h from 06:00 to 22:00 h daily. Results show that NAL precipitated intense somatic signs of withdrawal when gating devices were ON. This effect was almost abolished when gating devices were OFF. BMI rapidly decreased after the gating devices were turned OFF with maximum weight loss occuring 12 h post-OFF position, and gradually returning to baseline values after gating devices were turned back ON. These results demonstrate the validity of the "ON-OFF" gating device for non-invasively and repeatedly inducing physical dependence to opiates over a prolonged time.