The number of secretory cells in the mammary gland is often cited as a major determinant of milk production. However, literature data for proxy measures of secretory cell number do not fully support such a claim. In particular, measurements of total mammary DNA in livestock explain only <25% of variation in milk yield, probably because of tissue heterogeneity for DNA concentration. Relative to BW, measurements of udder size in dairy cattle, as total DNA or udder weight, are approximately double those seen in most other livestock classes. Therefore, selection for dairy production, not surprisingly, has resulted in cows with greater secretory capacity. There is limited evidence that genetic selection is still increasing udder size in some cattle populations, but more recent data are needed. It is contended that the most important period of mammary growth for determination of milk yield is that occurring in pregnancy and early lactation. Mammary development is largely complete, at term, in sheep, goats, and cattle, but in pigs, the udder continues to grow during the first 3 wk of lactation, depending, in part, on litter size. Increased litter size in sheep and goats will enhance the extent of mammary development at the end of gestation (and milk yield) by 20 to 25%. However, twinning in dairy cattle does not affect milk production and, by inference only, is not likely to affect numbers of secretory cells at term. Milking frequency and suckling intensity in very early lactation will increase milk yield in cows and increase mammogenesis and milk yield in sheep, indicating that even at term, the ruminant gland retains some capacity for further development, if demand requires it. There is limited understanding of the hormonal signals in pregnancy that regulate mammary development relative to the number of young carried. Furthermore, the genetic differences between dairy and beef cattle that lead to substantially greater udder size in the dairy breeds have not been identified. During lactation, the drivers for secretory cell loss in relation to milking frequency and nutritional status are still not known. Measurement of mammary development and using this phenotype in genomewide association studies to identify key genetic variants for mammogenesis will provide knowledge that is fundamental to understanding the quantitative regulation of milk production.