Fluctuations in reproductive performance (i.e., spawning success, hatching rate, larval survival) are a common occurrence in abalone breeding programs, in particular during the early stages of their development. Such fluctuations affect the numbers of families available for progeny testing and selection, and can have consequences for genetic gains and inbreeding. We used stochastic computer simulations to understand how genetic gains and levels of inbreeding are affected when greenlip (Haliotis laevigata) breeding programs encounter varying severity and frequency of reproductive failure. We simulated breeding programs for greenlip abalone with both conservative and aggressive selection approaches over 35 y (10 generations). Without reproductive failure, genetic improvements of 36%–55% could be achieved after 10 y of selection in a single trait in a commercial abalone breeding program with a conservative selection approach, and gains of twice that could be achieved with a selection approach that allowed high rates of inbreeding. A conservative selection approach would be sustainable even at high rates of reproductive failure, whereas a more aggressive approach would lead to nearly twice the recommended level of inbreeding. It was concluded that breeding programs for greenlip abalone may be buffered against unexpected fluctuations in reproductive performance if the selection approach is chosen strategically.