I agree that negative controls should be run on each IHC block.  However, I
don't agree that they should be run at the same 'dilution' as the primary
antibody.  Negative control serum or control ascites fluid is oftentimes
supplied at a much higher protein concentration as primary antibodies.
Check your antibody data sheets.  For primary polyclonal antibodies, find
out what the protein concentration of the stock solution is (usually stated
as ug/ml  or mg/L), then calculate what that would be once your antibody is
diluted out to its working dilution (usually somewhere between
0.1ug/ml-10ug/ml).  Then go to your negative rabbit serum and find out its
protein concentration.  Calculate what dilution you'd have to mix up to
'equal the protein concentration of the primary antibody working dilution'.
This can also be done with control mouse ascites fluid used as the negative
control for monoclonal antibodies.

If you are staining with ready-to-use antibodies (thus not figuring out
primary antibody dilutions), then I would suggest picking a mid-range
concentration of negative control serum/ascites fluid, such as one having a
final protein concentration of around 1ug/ml.

I've bought NRS in the past that has a 20X higher protein concentration than
my primary antibody, so if I were to use it at the same 'dilution' as the
primary antibody, there would be 20X higher nonspecific background.

I know this takes a lot more math, but it's much more accurate in giving you
a true picture of what nonspecific binding is due solely to host species
serum/ascites constituents at that concentration of antibody.

Jan Shivers
U of MN Vet Diag Lab