Cells containing pathogenic mutations in mitochondrial DNA (mtDNA) generally also contain the wild-type mtDNA, a condition called heteroplasmy. The amount of mutant mtDNA in a cell, called the heteroplasmy level, is an important factor in determining the amount of mitochondrial dysfunction and therefore the disease severity. mtDNA is inherited maternally, and there are large random shifts in heteroplasmy level between mother and offspring. Understanding the distribution in heteroplasmy levels across a group of offspring is an important step in understanding the inheritance of diseases caused by mtDNA mutations. Previously, our understanding of the heteroplasmy distribution has been limited to just the mean and variance of the distribution. Here we give equations, adapted from the work of Kimura on random genetic drift, for the full mtDNA heteroplasmy distribution. We describe how to use the Kimura distribution in mitochondrial genetics, and we test the Kimura distribution against human, mouse, and Drosophila data sets.