This study compares population-wide positive assortative mating (PAM) with open-nucleus breeding with an elite and main population when more effort is allocated to parents of the elite. A companion study showed that PAM is advantageous when testing effort is independent of parental value. In the present study,unbalanced testing was imposed by varying the number of crosses or the number of genotypes per cross. These unbalanced alternatives are compared with PAM, where the testing effort was varied so that better parents were mated more frequently. More effort allocated to parents of higher rank increased the additive effect and the additive variance and only slightly altered the group coancestry and inbreeding in the breeding population (BP) compared with completely balanced scenarios. Of particular interest to the breeder, large enhancement of the additive variance in the BP contributed to higher gains in the production population (PP). These simulations demonstrate that population-wide PAM leads to higher genetic gains compared with open-nucleus alternatives at any desired target level of diversity in the PP. This is true for both balanced (part I)and unbalanced distribution of testing effort (part II).