Hi Alok,
Physically, yes, the ADC is reading at that resolution. But there are many other factors in the surrounding circuitry which will effect sensitivity and accuracy. Simply dividing a 20V range by (2^24)-1 is the ideal textbook accuracy able to be achieved. There's some great information regarding the differences of resolution, sensitivity, and accuracy in our DevZone. As the 4461's accuracy on the range you are interested in should be around 280mV, .3mV of noise is a very good reading.
Your application, are you trying to read the accuracy of a DC signal? If so I would recommend either using one of our DMM products or averaging your samples from your DSA board. It sounds like you may have the wrong tool for the job but I'll have to have some more information to be able to recommend a product to you. Some external circuitry may need to be implemented to amplify the noise from your source to be read on your board. Please note that the DSA board is meant to study dynamic signals and is not the preferred tool for measuring accuracy of a static signal. These devices are very precise and thus measuring the characteristics of change in a signal are there intended purpose. Again, take a look at the following KnowledgeBase to understand accuracy with a DSA device.
Absolute Accuracy of Dynamic Signal Acquisition Devices
http://digital.ni.com/public.nsf/allkb/BA704FDCBB6C9C4E86256FAC006DB66B?OpenDocumentUnderstanding Instrument Specifications -- How to Make Sense Out of the Jargon
http://zone.ni.com/devzone/cda/tut/p/id/4439NI 446x Specifications
http://www.ni.com/pdf/manuals/373770h.pdfIncrease Your Measurement Accuracy by 10X with Signal Conditioning
http://zone.ni.com/devzone/cda/tut/p/id/3493World’s Most Accurate Multifunction Data Acquisition Device
http://zone.ni.com/devzone/cda/tut/p/id/4869
PBear
NI RF