When offered a choice, monkeys will choose cocaine over food. Or at least that's what three rhesus monkeys did during an eight-day study at the Medical College of Virginia in Richmond.

The study was reported in the Aug. 11 issue of Science, the official journal of the American Academy for the Advancement of Science.

The monkeys chose high levels of cocaine and ate almost nothing, according to the report authored by Thomas G. Aigner, then a graduate student, and Dr. Robert L. Balster of the college's department of pharmacology.

One general conclusion to be drawn from the study, Balster said yesterday, is that monkeys, capable of self-administering drugs, make good models for laboratory tests of some aspects of human drug use.

According to the Science report, three monkeys with "prior experiemental and drug histories" were fitted with venous catheters and individually housed so that they could choose every 15 minutes between five one-gram food pellets or an injection of 0.3 milligrams of cocaine hydrochloride per kilogram of body weight. The dose was within the range used by humans.

The monkeys, who also could choose to do nothing, were trained to execute a series of maneuvers involving lights and levers in order to turn on a green light, which resulted in food, or a red light, which signaled cocaine.

They were not fed for 23 hours prior to the start of the study and during it were given only the food they earned.

For the first two days, the monkeys chose cocaine at nearly the maximum possible level, missing only when they slept for brief periods. They ate no food.

"In an attempt to force the animals to make food choices, the light colors were reserved" after 48 hours for two monkeys and after 72 hours for the other.

The animals, expecting cocaine, got food. They ate very little of it and went to sleep for six to 12 hours.

They awoke and began experimenting with their light systems until they learned that the lights had been switched.

From that point on the monkeys again concentrated on cocaine, though during the last four days they were less regular users, mostly because they were sleeping for short periods of time.

Throughout the study, "the drug was almost exclusively chosen . . ." Even following periods of no choice behavior the animals did not attempt to obtain food. This exclusive preference for cocaine persisted for eight days." the report stated. "Concern for the health of the animals prohibited extending the testing period."

Balster said that a further conclusion stemming from the study contradicts past attempts to "explain drug use through pathology - that drugs are used to escape trouble or avoid reality.

"My view is that it makes them feel good and the animal tests confirm this. You ask people why they take cocaine and they say because it makes them feel good. It's time we start believing that," Balster said.