The purpose of this study is to evaluate the cost-income and genetic progress in a simulated buffalo population by implementing classical and genomic selection programs with different ratios of artificial insemination. The target buffalo population was simulated based on the progressive process in 10 replications and a genome consisting of 4 chromosomes with 2500 markers and 100 QTLs, using QMSim software. A population with an effective size of 1500 animals was created in order to make a linkage disequilibrium between markers and QTLs. From the last generation of the base population, two populations were considered: buffalo cows with artificial insemination and those with natural mating. Five evaluation and selection methods were examined for 5 generations: progeny testing with 100% artificial insemination, 20% milk recording and 4 superior buffalo and genomic selection with 20, 50, 80 and 100% artificial insemination and 4, 5, 6 and 7 superior bulls. Comparing two genomic and classical evaluations with 100% artificial insemination, the results showed that genomic evaluation can enhance the genetic progress and also lower the average of inbreeding in the studied buffalo population. Also, comparing the cost of genomic evaluation programs with different levels of artificial insemination showed that increasing the level of artificial insemination elevated the program costs. The highest cost and income were obtained for the genomic selection program with 100% artificial insemination. Results showed that despite the high investment and maintenance cost, genomic evaluation is more economical in terms of profit than classical evaluation in the selection of buffalo bulls.