An improvement in sugar content of sugarcane increases sugar yields with only a small marginal increase in costs of
production. This makes gains in sugar content economically more beneficial than corresponding increases in cane yield, and
means that increased sugar content is an important objective of sugarcane breeding programs. However, comparisons of cultivars
released in different years indicate that sugarcane breeding programs have delivered increased sugar yields via improvements in
cane yield, with much smaller contributions from sugar content. This is contrary to what might be expected, given that sugar
content normally has moderate to high heritability and is not substantially affected by competition effects in small plots, which
should make for easy gains from selection. Possible reasons for slow rates of genetic gain in sugar content include: that
insufficient weighting has been applied to sugar content in comparison with cane yield in selection of parents, that most
favourable alleles for sugar content are fixed in current cultivars, and that gene effects contributing to levels of sugar content
above current cultivars are negatively correlated with cane yield. Ways to test these hypotheses and address the associated
limitations within sugar cane breeding programs are proposed.