I'm using a pretty simple integration file for my DRO analysis...
Running Chemstation D.02.01...
Initial Area Reject 0
Initial Peak Width 0.040
Shoulder Detection OFF
Initial Threshold 10
Time[min] Integration Event Name
0.200 Baseline Now
0.200 Baseline Hold ON
0.200 Negataive Peak ON
0.200 Solvernt Peak ON
2.420 Solvent Peak OFF
3.200 Integrator ON
3.200 Area Sum ON
13.00 Area Sum OFF
13.00 Integrator OFF

Now I realize that invoking "Negative Peak ON" negates the solvent Peak and baseline hold commands.

So, if I have both "Baseline Hold ON" and "Negative Peak ON" I get a baseline that works very well for the upper end of the calibration but has enough excess area that the lower end of the calibration is biased high. This version also works very well for samples that have a rising baseline due to heavy oil range contamination that affects the high temperature end of the DRO range.

If I drop "Baseline Hold ON" I get the same good fit overall but a better fit in the low end of the detection range. However, its at the expense of not quite handling a rising baseline due to heavy contaminants. For example M, MD with ~350 ppm heavy DRO and an MS, MSD reading only 950ppm instead of 1300 ppm for a 1000 ppm spike. If I take out the "Negative Peak ON" command, I get M, MD of ~350ppm and MS, MSD of ~1300ppm.

My lowest standard is 50 ppm and I set that as my RDL. My 50 ppm level is 47ppm with Negative Peak ON and 62 ppm with Baseline Hold ON. Solvent blanks give me 9-11 with Negative Peak on and 18-22 with Baseline Hold ON.
Ottowa Sand method blanks rise from ~30ppm to ~60ppm because of the baseline hold. I really do need to soxlet extract that sand.

So, is using "Baseline Hold ON", the best way to fit DRO or am I missing a clue?

Edit: I had to fix my wording. Dropping "Negative Peak ON" makes it completely messed up. Apparently setting the baseline before the solvent peak makes "Baseline Hold" go nuts.