5% non-fat dried milk might be a good choice - like you have mentioned - and worth a try.  But, interestingly, immunity against BSA is a general concern due to dietary and medical treatments in humans (see the following link):

http://www.sciencedirect.com/science/article/pii/S0022175905000761

And so... given this, non-fat dry milk may also give you an unwanted bovine lactoglobulin response as well.

Fish gelatin then would be a good choice for this (smells really bad; but does work very well)...

I would suggest that human albumin or antibody negative serum could be used, as long as cost isnt prohibative

From the sentence that you mentioned that PBST would cause unused wells to be no longer coated, could you tell me what does PBST do to the wells? because I'm also having the same problem.

Thanks!

Perhaps the surfactant nature/action of Tween 20 interrupts protein attraction/sticking/coating of the positively-charged plastic surface of the wells...

BSA causes background a little bit. It probably reacts with antibody, depend on a grade of BSA. However, it is expensive.   Immunoblot test also shows background with BSA-blocking reagent.   I have used (1-2%) skim milk for blocking step both ELISA and blot. Gelatin is good for ELISA and blot but have to melt before use.   Also, casein is a good blocking reagent.  Skim milk is cheaper.  Here, all gives low background, except BSA.

You could try 5% non-fat skimmed milk or 1% Fish Gelatin. These have worked well in my hands.

Indeed many meat eating persons will have anti-BSA antibodies.  Your test serum and your enzyme conjugate should all be diluted in BSA containing buffer, if the wells are blocked with BSA.  Same will be if milk or casein will be used for blocking: always add the blocker to the antibody diluent.

Try 10% NFDM in PBS-Tween-0.05%, it works for me.

Regarding your question about the possibility of using PBS-T20 as blocking reagent, the answer is yes... I used 1%T-20 in PBS to block and 0.1- 0.05% T-20 in PBS for washing and as assay diluent.

Best

Forget all this blocking voodoo. It's needed in commercial assays (where the tests should be storable at room temperature for 18 month or more).

After coating (0.1 mol/L carbonate buffer pH 9.6, 2 ... 10 µg/mL antigen/antibody concentration) remove as much as possible of the fluid and store the plates frozen upside down. Blocking with a detergent containing buffer before freezing creates also problems: the detergent will go slowly but surely between your protein at the solid phase and the plastic surface.

Before using you have to wash your plates with PBS 0.1% Tween (or any other buffer at physiological pH and some detergent). The detergent has two parts, a hydrophilic directed into the water and a hydrophobic. This part will bind immediately at the hydrophobic plastic surface of the microtiter plates. Any other step is not needed and will create only trouble. Using milk proteins will give you high background in samples from patients with a milk allergy, BSA is both pure and expansive or don’t use it. Allergies vs. bovine proteins are also possible.