Sausages were produced with seven different mixtures with skim milk po
wder, sodium caseinate and whey protein (1.5, 3.5%) according to a sim
plex-centroid design, where the proportion of each ingredient varied f
rom 0 to 100%. Principal component analysis (PCA) identified that saus
ages with 1.5% milk protein were most similar in sensory quality to th
e controls and had minimum cooking loss. PCA was effective to reduce t
he number of attributes to five to describe the main variation among t
he 1.5% milk protein sausages. A mixture of 1:1 blend of skim milk pow
der and whey protein resulted in the product with lowest cooking loss.