letters to nature 434 NATURE | VOL 403 | 27 JANUARY 2000 | www.nature.com ®xed to the skull using dental acrylic. About a week after surgery, animals were implanted with morphine pellets and behavioural testing began 4 days later. Infusions of 0.5 ml per side were made through 28-gauge injector cannulae over 1 min, and cannulae were left in place for 1 min. All drugs were made fresh each day and were dissolved in ACSF. Dye was injected after the experiment to mark the injection site in all animals. Conditioned place-aversion procedureA balanced place-conditioning procedure was used to measure aversion in a chamber with two distinct sides 10 . On the ®rst day (preconditioning day), rats were allowed free access to both sides of the chamber for 15 min. Animals that spent more than 80% of the time on one side were eliminated. On the next two days (pairing days), the animals were given an intraperitoneal injection of naltrexone (1 mg per kg) or saline, and were con®ned to one side for 30 min. Animals given naltrexone on pairing day 1 were given saline on pairing day 2 and con®ned to the opposite side, and vice versa. All adrenergic drugs were microinjected on each of the two pairing days 5 min before naltrexone or saline; controls were similarly injected with ACSF. During pairing, an observer scored each occurrence of somatic withdrawal signs. On day 4 (test day), animals were given no drug injections and were returned to the test apparatus for 15 min with free access to both compartments, and the time spent in each compartment was measured. For shock training, place conditioning was carried out in drug-naive animals as above, except that, on pairing day 1, animals received a 0.8 mA foot shock (randomly given for 1 s every 3 min through the chamber oor) over the course of the 30-min session; on pairing day 2, they received no foot shock and were con®ned to the opposite side.