ND filters are a good way to test. Usually the problem is that the meter becomes unreliable at very low light levels.

You can see this with the meter in a 35mm camera--try seeing if you get the same reading with the lens wide open as you get in stopped down mode with the lens at f:22. In most cases, the readings will be different. It's not as much of a problem with handheld meters, but if you think you're getting questionable readings, then it's worth testing.