The objectives were (i) to evaluate the effect of temperament, determined by modified 2-point chute exit and gait score, on artificial insemination (AI) pregnancy rates in beef heifers following fixed time AI and (ii) to determine the effect of temperament on cortisol, substance-P, prolactin and progesterone at initiation of synchronization and at the time of AI. Angus beef heifers (n = 967) at eight locations were included in this study. At the initiation of synchronization (Day 0 = initiation of synchronization), all heifers received a body condition score (BCS), and temperament score (0 = calm; slow exit and walk or 1 = excitable; fast exit or jump or trot or run). Blood samples were collected from a sub-population of heifers (n = 86) at both synchronization initiation and the time of AI to determine the differences in serum progesterone, cortisol, prolactin and substance-P concentrations between temperament groups. Heifers were synchronized with 5-day CO-Synch+ controlled internal drug release (CIDR) protocol and were inseminated at 56 h after CIDR removal. Heifers were examined for pregnancy by ultrasound 70 days after AI to determine AI pregnancy. Controlling for synchronization treatment (p = 0.03), facility design (p = 0.05), and cattle handling facility design by temperament score interaction (p = 0.02), the AI pregnancy differed between heifers with excitable and calm temperament (51.9% vs 60.3%; p = 0.01). The alley-way with acute bends and turns, and long straight alley-way had lower AI pregnancy rate than did the semicircular alley-way (53.5%, 56.3% and 67.0% respectively; p = 0.05). The serum hormone concentrations differed significantly between different types of cattle handling facility (p < 0.05). The cattle handling facility design by temperament group interactions significantly influenced progesterone (p = 0.01), cortisol (p = 0.01), prolactin (p = 0.02) and substance-P (p = 0.04) both at the initiation of synchronization and at the time of AI. Inter- and intra-rater agreement for temperament scoring were moderate and good (Kappa = 0.596 ± 0.07 and 0.797 ± 0.11) respectively. The predictive value for calm and pregnant to AI was 0.87, and excited and non-pregnant to AI was 0.76. In conclusion, the modified 2-point temperament scoring method can be used to identify heifers with excitable temperament. Heifers with excitable temperament had lower AI pregnancy. Further, cattle handling facility design influenced the temperament and AI pregnancy.