I am assuming you have some knowledge of isocratic type elutions, where you have a constant mobile phase throughout your anaylsis.
In isocratic methods, a higher percent organic content (referred to as %B from here on out) will lead to sooner elution time. Likewise, if you decrease the %B, you will have a longer elution time.
Gradient elution is pretty similar in some ways. Instead of changing %B though, you change your %B per minute. For instance, if you increase your %B/min, your peaks will elute sooner. If you decrease your %B per minute, your peaks will elute later. Higher %B/min gradients lead to sharper peaks, but will also push any peaks you have closer together.
One important thing to note is that every HPLC system has a "void volume". In an isocratic method, this isn't as important as in gradient methods. You can typically see the period of time it takes for the void volume to be run through by a dip in the baseline of your chromatogram. This is the time that the injection starts to get to your detector. Data handling systems typically record your chromatogram starting at the time your sample is injected in the system.
Let's say that the time for the void volume to pass is 2 minutes. That means the initial conditions of your gradient are hitting the detector just then. Even if you have your instrument starting at 0%B and going at 10%B/min, that portion of the gradient isn't hitting the detector until 2 minutes because of the 2min void volume. With this knowledge, you can estimate approximately what %B range your peaks are eluting. An example of this is, assuming you have a 2 minute void volume and a gradient that is 10%B per minute, if you have a peak eluting from 4-5 minutes, that peak is actually being eluted at 20-30%B because the 2 minute void volume does not have the gradient running through it.
IF you determine that your peaks of interest are coming out from, for instance, 30%B to 50%B you can clip your gradient to run from perhaps 20% to 60% so your gradient is not as long.
In simpler terms for gradients:
Adjusting %B/min higher will lead to quicker eluting, sharper peaks but could lead to bad resolution between peaks.
Adjusting %B/min lower will lead to slower eluting, less sharp peaks but could lead to better resolution between peaks.
Determine your void volume visually if possible. This is the point where your gradient is actually starting (if your instrument method does not have an isocratic portion at the beginning). This can help you to know at about what %B your compounds are eluting.
This is a pretty basic view of gradient type elutions. I am not completely sure of any on line resources and will let others add to this.
. See if you can get your hands on one of the following: