When we say that something is a genetic marker we mean that we can establish its linkage to a chromosome AND that we have some way of discerning, or detecting, how the marker has segregated after meiotic recombination (this definition is only valid for diploid species that undergo sexual reproduction).
So, for example, in the fruit fly, Drosphila melanogaster, the white gene is linked to the X chromosome. Wild-type flies have red eyes, a mutation in the white gene yields flies with white eyes. You can use genetic crosses (matings) to (a) show that the white gene is linked to only one of the four fly chromosomes, and (b) if you have other genetic markers on that chromosome you might be able to construct a genetic map showing the order and distance of the linked genes.
You can also use a DNA fragment from a chromosome as a genetic marker--if you have an assay that lets you track the fragment after meiotic recombination. A Southern blot, using a labeled probe is one way to accomplish this (e.g. a Restriction Fragment Length Polymorphism or RFLP). PCR that lets you detect a RFLP or a SNP serves the same function, and is much quicker than a Southern Blot (and requires less DNA starting material).
A QTL is just one of these "molecular" genetic markers that shows linkage to the trait you are assaying. So, for example, if you had RFLPs or SNPs sprinkled all along the human chromosomes, and if you had a way to detect the segregation of all those markers after meiotic recombination, then you could use DNA from a family with lots of kids of varying heights, where one parent was tall, and the other parent was short, to see if any of your genetic markers were segregating, or linked to the phenotype of height in the children (this example assumes that height in humans is a quantitative trait).