Well, the 0.025% part means the multimeter reading will be accurate to within 0.025%. I think the number that follows is the number of counts - it's to do with rounding error and noise limitations.
I would guess that the DVM has a 5 digit display (or 4 3/4 digit ie 399.99).
As for the +/- 0.025%, this is reasonably obvious, except that it does not state whether that is 0.025% of the actual reading, or 0.025% of full scale.
The +5, I assume this means that the last digit of the display could be up to 5 bits out.
So, if the meter were reading a standard 100.000 volts, it could display anything between 99.9 and 100.15 volts.
(400v range 0.025% = 0.1 volt, and add 5 bits to the last digit).
Accuracy specifications can be very frustrating to figure out. Manufactures are of course in competition with others so such specs are important to their marketing efforts and not always totally honest or at least misleading.
Accuracy is not a simple specification where one number can totally describe the meters attributes and most don't publish their full testing protocol.
Typically you get what you pay for in this area and their is a reason the Fluke and HP (agilent) meters are so respected by both the pros and hobbyist.
I would guess that the DVM has a 5 digit display (or 4 3/4 digit ie 399.99).
As for the +/- 0.025%, this is reasonably obvious, except that it does not state whether that is 0.025% of the actual reading, or 0.025% of full scale.
but how about the "+30"?
AC Voltage 4V/40V/400V/1000V (0.4%+30)
Do you think this model UT71E of (China made) DMM is good?
I have seen the actual thing, it look quite solid.
It can link to PC through USB or RS232 too...