a) Select desirable plants

• Number depends on variation of original population, space and resources for following year progeny tests
• Selecting too few plants may risk losing superior genetic variation
• A genotype missed early is lost forever

b) Seed from each selection is harvested individually

c) Single plant progeny rows grown out

• Evaluate for desirable traits and uniformity
• Should use severe selection criteria (rogue out all poor, unpromising and variable progenies)

d) Selected progenies are harvested individually

e) In subsequent years, run replicated yield trials with selection of highest yielding plants

f) After 4-6 rounds, highest yielding plant is put forward as a new cultivar