like the others said, it can not be done.
The main problem I would see is that the transistor would oscillate even if you could get a reading with a multimeter--rendering that reading meaningless.
If you need to test it, I would recommend making a microwave test fixture. The gate and drain would need DC blocking series capacitors and 50 ohm terminations to ground. There would be a terminal on the test fixture for gate bias, and another one for drain bias. The gate bias structure would be properly designed using S parameter technique to not cause an oscillation, and to only let very low frequencies out that bias terminal. Similarly, the drain bias structure would do the same, but probably have perhaps a 200 ohm series resistor in it (to limit current). The device will need excellent grounds from the source to ground with vias close to the package.
After building such a test fixture, you would hook up the drain to a standard dc power supply with a current meter. You would hook up a similar power supply to the gate bias network. Varying the gate voltage should result in the drain current going from zero to some maximu level. That confirms it is a working device.
Do NOT exceed the voltage and current maximums for the device! Otherwise it will be a very short test