Though it will not help bleed-through problem, you will certainly benefit from the using the  labeled secondary antibodies which are pre-absorbed.

I tend to use the highly cross-absorbed, but if the signal-to-noise of the primary antibody is good, and you have a good confocal, in most of cases you can go with the standard secondary abs....

Hmm, I have not really seen this. Is it the secondaries or is it the filters and dyes you use?

I prefer Vector Laboratories reagents. I cannot recall to have had problems with any of their secondary antibodies, streptavidin conjugates etc. I don't like their mounting medias not so much anymore. There seems to be too much DAPI in some of them.

Your secondary system should never be in doubt. The primary antibodies are enough of a source of problems. If you introduce another layer of problem than things get too complicated. If there is bleed through you should get a better system. What do you currently use?

The only problem I ran into has been related to using anti-IgG1, -IgG3, -IgM and other isotypes. They often do not so well. Unfortunately, Vector Labs does not really offer choices in this field and there might be a reason why.

I've switched to using secondary antibodies screened for species cross-reactivity when doing multi-antigen staining. Though they bind less potently, cross-reactivity is a real concern. For a while I had false positives of peptide colocalization because an anti-goat secondary was binding to a rabbit primary. After switching to a different anti-goat secondary with minimized cross-reactivity, the problem disappeared. However, keep those non-specific secondaries around. If you're not doing multiple-antigen staining, they bind much better to the primary antibodies.

Is the "bleed through" due to the secondaries, or the confocal? Can you adjust the spectral filter settings and narrow the excitation/emission bands, or are these fixed by filter cubes in your confocal system? Is the pinhole set correctly? Are you scanning the different colour channels sequentially rather than simultaneously? For the secondaries, I would say use the highly cross-adsorbed, so as to reduce any cross-reactivity. Another option is to try reducing the concentration, as intense signals may "shine through" a filter into a longer wavelength and excite it. I often label with streptavidin-conjugated dye, and the intense signal is sometime better in the infra-red, so it doesn't shine through into the green or red channels. Try a control with all primaries and just one secondary, and see which channels are visible...

Personally, I have never problems with non specific binding with secondary antibodies. You have to experiment each secondary antibody alone to verify it. After, if there is no problem, try to use them in multiple labeling. At this step, you can verify if you have not cross reactivity. Good luck.

After a review of our controls and the direction of the apparent bleed-through, I've concluded that our secondaries really aren't the problem. Since it's our weakest primary that's "shining-through" to our other primary and diluting that first secondary isn't really an option, we're going to expand our confocal range with a new cube and hopefully get the second primary out of the emission tail of the first. We're also taking extra care to avoid having to use two primaries from the same host. 

Thank you everyone for your input! 

Dear Cynthia,

your original question did mention nothing about using 2 primaries from the same species. If you have a question, it is important to reveal some key details. Using for example 2 mouse primaries is most of the time tricky and then definitely a secondary problem. Therefore, I really don't understand your explanation.Honestly, it makes no sense.

Dear Cynthia,

As Thomas says above, two primaries raised in the same host will give you overlap. Changing your confocal system will not help. The secondaries are raised against common Ig parts of the host the primary, and have nothing to do with the specificity of the primary for it's antigen target. Even if you apply your primaries sequentially, the secondary for the 2nd primary will also recognize and bind to the 1st primary, possibly displacing some of the first 2ndry. Thus your overlap is probably unavoidable, even if you try something tricky like blocking the primary host Ig sites after you label it with it's appropriate secondary. Best to try a different primary antibody - even rat and mouse can be separately detected, although you must check for cross reactions.

My apologies - I shouldn't have mentioned the duplicate-host primaries question. We were having the overlap problem even in primaries from two separate hosts. The addition of a far-red filter cube and the deletion of  a particularly weak primary antibody seems to have fixed the issue. Thank you all for your help!