Genes, stored in DNA, are usually copied faithfully from parent to offspring. Muller exposed fruit flies to X-rays and showed quantitatively that many flies developed sudden changes in eye color or wing shape—mutations. He also discovered a dose-response relationship: the more X-ray energy absorbed, the more mutations appeared. This proved that radiation is a powerful mutagen, shaping safety standards for nuclear technology and medical X-rays and opening the door to "radiation breeding" for new crop varieties.

Muller employed 50–90 kV X-ray generators and devised the ‘cross-test’ using recessive lethal and visible markers on the X chromosome. By irradiating male flies with defined doses and scoring mutation frequencies in F1 and F2 generations, he demonstrated induction rates hundreds of times above spontaneous background and an approximately linear dose-response. He proposed that radiation physically breaks chromosome structures, causing deletions, substitutions, and inversions. Muller's work laid the foundation for radiation genetics, cancer etiology studies, and debates on molecular evolutionary rates, and it was later incorporated into risk models by the International Commission on Radiological Protection (ICRP).

Muller’s X-ray mutagenesis paradigm utilized the sex-linked recessive lethal assay in D. melanogaster, establishing a quantitative LNT (linear-no-threshold) model for germ-line mutation rates. Dose ranges (0.5–40 Gy) revealed a near-linear increase in induced single-gene and multi-locus events, interpreted as double-strand breaks followed by mis-repair. He distinguished segmental aneuploidies from point mutations via cytological analyses of giant salivary-gland chromosomes. The work precipitated intensive studies on radiogenic clastogenesis, influenced the BEAR I & II genetics panels, and remains a cornerstone in contemporary discussions of low-dose stochastic effects and mutational spectra in radiation oncology and space biology.