I have seen some Van Deemter's curves (HETP vs flow rate) which show that optimum flow for best resolution should be ~1.3 mL/min for a 5 µm particle size, 4.6 mm id column, 65% ACN/water. Other plots (N or 1/N vs flow rate) and theoretical evaluations show that maximum plate count N occurs for lower flows of 0.4-0.7 mL/min for 50% or 70% ACN/water. Is there an optimum flow rate for maximum resolution for 5 µm 4.6 mm ID reverse phase columns or does it vary greatly depending on mobile phase and analyte?

There is an optimum flow rate for high performance concerning plate high but it isn't hint for high resolution.So you focus on the issue when other conditions are definited.

It varies greatly depending on mobile phase, analyte (molecular weight), temperature, stationary phase pore size distribution (and probably the phase of the moon).

However, remember that resolution depends on the square root of the plate number, so that it takes a large change in N to make much of a difference. The rule of thumb for small molecules and column packings of 5 micron and below is to run as high a flow rate as your pressure limits (what you are comfortable with) will allow and focus on selectivity to improve resolution. Then, as the previous post said, you can adjust flow rate to tweak N when the other conditions have been pinned down.

If you deal with big molecules, life is a bit more complicated; flow rate can make a big difference.

The reason for this posting is that I have a method that is run with 55% ACN/water with a 4.6 x 250 mm, C18, 5 µm column at 1 mL/min. Would running at 0.5 or 0.7 mL/min be expected to improve resolution?

We normally run 1.00 ml/min on 4.6 mm columns, and 0.25 ml/min on 2.1 mm columns (used most). Slowing down the flow rate a little might improve resolution a little, but so would changing to a 3u particle size. I would say that changing to a smaller particle size would not require any re-validations (maybe minor or POV) because you wouldn't be changing the chemistry of the separation.

Under the conditions that you gave, I expect you to have about 15 000 plates for a standard small molecule at room temperature. Going to a slower flow rate, you may get about 21 000 plates at around 0.3 mL/min. In my opinion, the gain in resolution (about 20%) is not worth waiting for (the analysis time will increase about 3-fold).