The inheritance of Fiji disease resistance in sugarcane was studied, s
o that strategies for breeding for resistance could be developed. Resi
stance was highly heritable and most of the genetic variance was addit
ive when infection pressure was low to moderate. Infection pressure wa
s very heavy in one experiment and this reduced discrimination between
parent clones and families. The data were skewed and this affected st
atistical and genetical analyses which showed that non-additive geneti
c variance was high. In general, however, it was concluded that select
ion of resistant parents would produce families with resistant progeny
. Some moderately susceptible families produced a small number of resi
stant progeny and, if these families have agronomic merit, the breedin
g strategy should enable such families to be planted. The computerised
strategy adopted rejects crosses if the average susceptibility of the
two parents is too high, and emphasises the use of more resistant par
ent clones. As some susceptible clones will be selected, clones are su
bjected to a field screening trial as soon as practicable. These strat
egies have reduced the percentage of susceptible clones in the populat
ion from 70-80% to 20-30%.