Numerical genetic changes can be most easily examined by simply preparing metaphase chromosomes and counting the number of chromosomes in the spread. Unfortunately, it is often impossible to obtain high-quality metaphase preparations from samples, especially solid tumors. Even more frustrating, it is just such tissues that are particularly interesting to study. For example, numerical imbalances in these tumors might identify the sites of either tumor suppressor genes in deleted regions or proto-oncogenes in amplified regions involved in the initiation or progression of that particular disease. A submicroscopic molecular method, such as loss of heterozygosity (LOH) or allelic imbalance (AI), will provide much greater detail in a small region, but is impractical to have that level of detail in a genome-wide screen. Because of these limitations, a novel methodology was needed to evaluate the numerical genetic composition of interesting samples, especially solid tumors.