Typical commercial bakeries in the United States are highly mechanized, mas
s-production facilities. U.S. hard wheat breeding programs use small-scale
physical dough testing and pup loaf bake procedures to identify and select
improved quality genotypes. The accuracy of such approaches in the predicti
on of commercial-scale quality performance is poorly understood. Samples fr
om six hard red winter wheat cultivars grown in 11 locations over three har
vest years were used to correlate grain hardness, small-scale test bakes, m
ixograph variables, and various measures of flour protein composition with
quality assessments from commercial test laboratories. Samples were milled
on both pilot- and small-scale mills. Protein content and 100-g pup loaf vo
lume were more often significantly correlated with commercial test bake var
iables than all other small-scale variables. Stepwise multiple regression m
odels explained on average, approximate to 40% of the variation in commerci
al test bake procedures. Mixograph properties, pup loaf volumes and absorpt
ion, and flour protein content were the most frequent variables identified
in model development. Pup loaf bake results an pilot- and small-scale mille
d flours were highly correlated. Differences in milling technology do not a
ppear to be a significant source: of error in relating small-scale test bak
es to commercial quality.