When I quantified my tau bands using ImageJ, I just boxed up both bands. If you do this, make sure you keep the box the exact same size for all lanes.

Thank you, Danielle. That's what I did as well, but then you get multiple peaks. Did you just add those values together? 

You should be able to draw a box around just the band of interest in each lane, without having multiple peaks.

I believe I drew a line to the left of the first peak, and a line to the right of the second peak, and used the entire area between the two lines (even the small area between the two peaks). This way there is less possible variation (human error) because you are only drawing two lines to identify both peaks together, instead of four lines to identify the two peaks individually. If you are consistent across all samples, this works fine. I did this for a paper (now published) a few years ago and haven't used ImageJ since, so I'm not sure if the software has been updated in a way that is better for area measurements  - if it has, the basic answer is yes, just add the peaks together.

Thanks Danielle! ImageJ seems to have just as much human error involved as ever, so drawing the line to connect both peaks sounds like the best bet. I appreciate the input. 

Hello Kristen,

Recently, I am also working on Tau and used tau5 antibody for my western blotting. My bands look more faint than yours. Could you share with me your detection protocol? What protein lysis buffer did you use? And what are the procedures for protein preparation? Thank you very much!

Grace