Protocol Online logo
Top : New Forum Archives (2009-): : ELISA and Immunoassay

Cut-Off Determination - (May/17/2011 )

Hi everyone,

I standardized an "in house" ELISA assay and this has been working fine. My doubt is about the Cut-off determination.
We have used this formule to determine the cut-off: mean Od from negative samples + 3* Standard Deviation from negative samples. The problem is half of ELISA plate is used to negative samples. This has worked fine but some authors has used this formule: mean Od from negative sample + 0.300 = Cut-off.
I would like to know what is the best way to calculate Cut-off? Would there be another formules?
What do you think ELISA's Expert?

Thank you in advance

Best Regards



This is a very conservative determination, which I have also used myself. Nowadays people usually go for the 5% false positive rate instead of an arbitrary cut-off. Either way, you don't have to determine this cut-off every single time you run the ELISA. If you "validate" your assay and prove that it behaves like it should every time you run it, you need only to run the standard curve and a positive and negative control in your experiment. The cut-off will hold for all ELISA plates. Not necessarily the signal value in OD (which is somewhat variable), but if you transform this into concentration units, these should hold. Assuming again that the interrun differences are low (e.g. CV less than 10%).