You can get valuable informations by following the link below:-

www.treat-nmd.eu/downloads/file/sops/dmd/.../DMD_M.1.2.006.pdf

You can further refer to the following link to have a look at the curve:-

http://www.biomedcentral.com/content/figures/1472-6750-12-51-1.jpg

What is the limit of detection of the assay? The amounts of hydroxyproline in your samples may have been too low. There is some confusion in your question between mass and concentration:

"I diluted a 1mg/mL hydroxyproline stock to 100ug/mL and then further diluted to obtain the following concentrations/well: 1 ug, 0.8ug,0.6 ug,0.4ug ,0.2ug , and 0 (blank with dH20). I aliquoted 10uL of each of these dilutions into a 96 well plate"

If you diluted to 1 µg/ml and below and added 10 µl, that is only

I have no idea of the limit of detection, but I am not sure of the Adam B Shapiro interpretation. His description is a double dilution. Anyway, I do not know the reason why you evaporate to dryness, but I suggest to avoid any heating. Hydroxyproline and Ehrlich reactive cannot be stable at those temperatures. 

Thank you for all of your responses! I increased the concentrations of my dilutions to 1mg/mL, 100ug/mL, 10ug/mL, 1ug/mL, and 01.ug/mL in different wells and that seemed to induce some color change (1mg/mL = red, 100ug/mL=orangish, and the rest were yellow like the Ehrlich's reagent). Unfortunately the curve is still not linear at all. I skipped the evaporation to dryness step that I was performing right before adding the Ehrich's/Chloramine T reagents , however, the sigma protocol calls for the samples to incubate at 60 C once the Ehrlich's and Chloramine T have been added to the hydroxyproline. Francisco Solano, if the are not stable at this temperature is there a different method you would suggest for the reaction to occur ?

To generate a standard curve, given the sensitivity you are seeing, use arithmetic dilutions between 0 and 1000 µg/ml, i.e. 0, 50,100, 150, etc. Your last standard curve may appear non-linear because you skipped the linear range by jumping from 100 to 1000 µg/ml. Even if the standard curve isn't linear, it can still be used by interpolation.