Answer: This comes down to personal preference but the main thing to consider is that for any single absorbance reading there are actually two or more components to that reading. This applies to all absorbance assays, and is not specific to this particular assay. An appreciation of the different components is required in order to determine the best way of handling the controls and blanks, and whether to subtract blanks from the assay wells and standard curve before carrying out calculations. For example, if we assume that the substrate is contaminated with free Pi, the single measured absorbance (Y1) for the assay wells is the sum of three separate components (i) the blank value due to the Gold reagent alone, which is ~0.1 (ii) the signal due to contaminating Pi (iii) the signal due to Pi released from the substrate during the assay. The control wells (in this case wells with substrate but without enzyme) give a single absorbance reading (Y2) that is made up of two components, the blank value and the signal due to contaminating Pi in the substrate. Thus subtraction of Y2 from Y1 subtracts the component due to contaminating Pi and also the blank component. The resulting value can therefore be used to calculate the amount of Pi formed using a blank-subtracted standard curve. Whilst in the above example it was not necessary to subtract the measured blank value directly from the assay data (since subtraction of the control Y2 from Y1 achieved the same result) it is generally safer to subtract the blank value (i.e. water plus 0.25 volumes of the reagent mix and 0.1 volume of stabiliser) from the standards, assay wells and any control wells before calculations on the data are performed. In this way, regardless of how many controls need to be subtracted from the assay data you cannot inadvertently subtract the ‘hidden’ blank value more than once. #Tip: you can do a single control that includes all assay components by using a different order of reagent addition. Add all components except the enzyme (but do not add water instead of enzyme) to triplicate wells, followed by the reagent mix (ignore the fact that the enzyme is missing and add the usual 0.25 volumes of reagent mix). Next, add the enzyme. (There are few, if any, enzymes that are active in the acidic medium). Five minutes later add the stabiliser and read the plates normally. This approach allows you to combine the enzyme and substrate in a single control well.