The objective of this paper was to compare 3 bovine AI protocols, evaluating the cost of each treatment and the pregnancy rates obtained, to demonstrate the importance of routine utilization of estrus synchronization techniques. The experiment was conducted in the Farm School of Cachoeiras de Macacu of the Federal Fluminense University, in the state of Rio de Janeiro, during the period from 2006 to 2007 (August to June). One hundred sixteen crossbred female Holstein-Zebu dairy cows in cycle, with body condition between 2.5 to 4 (on a scale from 1 to 5), were selected and divided randomly in 3 groups (G). Animals in G1 [n = 60; 21 heifers (H), 13 primiparous (P), and 26 multiparous (M)], with palpable corpus luteum, received 25 mg of prostaglandin F2 intravulvar, and AI was done 12 h after estrus detection. For animals in G2 (n = 30; 11 H, 6 P, and 13 M) an intravaginal progesterone device (CIDR®, Pfizer Animal Health, New York, NY, USA) was inserted and 2 mg of estradiol benzoate (EB) was administered i.m. on Day 0. On Day 8 the device was removed and 25 mg of prostaglandin was administered i.m. on Day 9, 2 mg of EB was administered, and 30 h later, timed AI of all animals was performed. For G3 (n = 26; 9 H, 5 P, and 12 M), the protocol used was similar to that for G2, except that on Day 8 400 IU of eCG was administered (in addition to the removal of the device) and on Day 9 the dose of EB was reduced to 1 mg and timed AI was done 27 h later. AI was done with semen from 2 sires with proven fertility, by the same technician. The pregnancy diagnoses were done on Day 45 by palpation per rectum and transrectal ultrasonography. The cost-benefit relation was calculated according to the cost of pharmaceuticals multiplied by the number of animals in each experimental group divided by the number of pregnancies obtained in each one, without considering the manual labor costs in all the groups and considering 3 uses of the CIDR. The statistical analyses were done using chi-square test. The pregnancy rate obtained in G2 (60%, 18/30) was superior to that in G1 (30%, 18/60) (P < 0.05), demonstrating the efficiency of the treatment that utilized progesterone. G3 (53.85%, 14/26) was similar (P > 0.05) to G2, despite the fact that the eCG was added in this group. In reference to cost-benefit, considering only the pharmaceuticals costs and the value for pregnancy obtained in each group, the costs found, respectively, were G1 U$7.02 and U$23.66; G2 U$28.57 and U$47.59; and G3 U$47.96 and U$89.05. The lowest cost in the estrus synchronization was based on the use of prostaglandins alone; the use of progesterone increased the synchronization cost, but also increased the pregnancy rate and eliminated the need for estrus observation. The addition of eCG in cows that are cycling does not necessarily demand a increase in pregnancy rate. Failure in estrus detection, using prostaglandin alone, increased the interval from calving, causing a daily loss in milk production, which justified the investment in protocols of estrus and ovulation synchronization, contributing to increase the reproductive efficiency of a dairy cow herd.