The LH peak in cattle is the most precise event for predicting ovulation beginning 24 h later, and thus, AI time. Previous studies demonstrated that embryo production was improved when AI was conducted 12 h before ovulation; that is, 12 h after LH peak. This study aimed to evaluate the benefit of LH peak monitoring with Predi′Bov® (ReproPharm®, Nouzilly, France) following superstimulation in order to optimize numbers of viable embryos (VE). Predi′Bov® is a rapid (40 min) and easy to use on-farm test allowing LH peak detection from a few drops of blood. The test was also used to estimate the variability in the time of the LH peak and onset of oestrus. This study was conducted by the embryo transfer teams of 3 French cooperatives, in collaboration with UNCEIA. Forty heifers in stations (Creavia, Midatest) and 23 cows on farms (GEN′Iatest) were superstimulated by 8 injections IM of Stimufol® or Pluset® (FSH1 to FSH8) over 4 days in 2011–2012. Donor station heifers were treated twice in a Latin square design with a reference protocol where AI was conducted 12 and 24 h after onset of oestrus, and in an experimental protocol where AI was conducted 12 and 24 h after a positive Predi′Bov® test. Semen of different sires was used for both protocols. The Predi′Bov® test was carried out on 3 blood samples (BS1,2,3) collected every 6 h beginning at FSH7 in stations and FSH8 on farms to detect the earliest LH peaks. To determine late LH peaks, Predi′Bov® test was carried out on BS4 collected 24 h after FSH8. Univariate statistical analysis was performed to look at the relationship between qualitative (chi-square) and quantitative (t-test) variables. The difference was considered significant when P < 0.05. The Predi′Bov® test showed that 37.5% (15/40) of LH peaks occurred during the last day of FSH treatment (BS1 or BS2) in stations and 26.1% (6/23) at BS1 on farms. At Creavia station (n = 24), the LH peak was detected anytime from 24 h before to 9 h after the onset of oestrus. In stations, the VE percentage did not differ whether AI was done following oestrus or LH peak detection (63.1% and 61.8% in reference and experimental protocols, respectively). In stations, the VE percentage from 9 females with an LH peak detected at FSH7 (BS1; 41%) in the reference protocol did not differ from the experimental protocol (50%). On farms, the VE percentage was numerically higher but not significant in the experimental protocol (65.4%, n = 16) compared to the reference protocol (47.2%, n = 7). Further investigations are needed, taking into account the effect of collection rank, sire, and female effects, to confirm the trends shown by these results. In conclusion, Predi′Bov® can be used as easily on farms as in stations. Its use allows the detection of animals that have early or late LH peaks, which in turn provides the opportunity of carrying out AI at the optimal time for such females.