Abstract The retailers are interested in different profit maximization approaches, which help them to perform daily shelf arrangement tasks, improve customers’ satisfaction and avoid out of stocks on the shelves. Most of the retail literature proposes basic models which don’t reflect complicated merchandising rules. The aim of the paper is to develop a shelf space allocation model which investigates vertical shelf levels, horizontal shelf division into segments, and product item allocation rules such as cappings and nestings. A genetic algorithm has been developed to implement the model. The efficiency was estimated with the help of CPLEX solver. The computational experiments show that the proposed approach allows getting the results of sufficient quality for different problem sizes in a short running time without requiring large computing resources. GA and CPLEX found statistically the same profitable solution within the same computational time. However, GA found a better solution in situations where the commercial solver can’t perform calculations during the increased computational time and stopped after a couple of minutes. This proves the reason for GA implementation for shelf space allocation problems.