A question close to my heart (me bein' only 5' 2").
Height is a multi-allele trait, but like many phenotypic traits it follows a bell curve distribution. So what Leolaia was saying about the range of heights found within our species is true. With the lucky Nordic gods at the higher end, and the pygmies, and yours truly , finding ourselves on the shorter end.
Pygmies and many of us shorter statured people have skeletal structures that aren't all that disproportionate in comparison to the average sized adult. Phenotypes we recognize as dwarfs though have disproportionate skeletal structures and thats where some unusual alleles are coming into play.
There are different types of dwarfism, so I don't know which one PeacefulPete had in mind when he gave some of the probabilities of dwarf offspring. If we take achondroplasia (a very prevalent form), the dwarfs would be heterozygous for the trait - one allele is "normal" and the other one (in this case dominant) confers dwarfism. Being homozygous - two copies of the dominant allele would be lethal. Crossing the two dwarfs: we'd expect 50% of the zygotes to be heterozygous (dwarfs), about 25% would not have any copy of the dominant allele (normal height) and 25% zygotes homozygous. So if we only consider the offspring that continue to live, the proportions would better resemble 2/3 dwarfs and 1/3 normal height.