Calibration of the Embryo Growth Model with Our Day 15 Data. The embryo growth model (Equations 1–5) was also fitted to data on the effect of circulating progesterone on embryo length on Day 15 (yr 1: farms 1–3; yr 2: farms 1 and 4). Progesterone on Day 15 is correlated with progesterone on Day 7 in yr 1 (R = 0.20; P = 0.02) and yr 2 (R = 0.47; P < 0.001). The ability of the model to predict embryo length on Day 15 is shown in Figure 2C for yr 1 (R2 = 0.08; P = 0.01) and yr 2 (R2 = 0.01; P = 0.4). The estimated model parameters for embryo growth are a = 0.00016 ± 0.00001 ng−1∙mL, n = 0.15 ± 0.002 in yr 1 (assuming g = 1, b = 4.75 d, A0 = 1, and h1 = 0.36 log10 cell number∙d−1) and a = 0.000025 ± 0.0002 ng−1∙mL, n = 0.14 ± 0.06 in yr 2 (assuming g = 1, b = 4.75 d, A0 = 1, and h1 = 0.40 log10 cell number∙d−1), which are significantly different than those estimated from our previous ET study (a = 0.0165 ± 0.001 ng−1∙mL and n = 0.466 ± 0.02, Shorten et al., 2018). If n = 0.15, then doubling progesterone decreases the doubling time by 10 ± 0.1%, whereas if n = 0.466 then doubling progesterone decreases the doubling time by 28 ± 1%. The weaker dependence of in vivo embryo growth on progesterone in this trial suggests that either in vitrotransferred embryos are 2 to 3 times more sensitive to progesterone or other factors had a larger effect on embryo growth in the current trial. The estimated effect of progesterone on the cell doubling time [log10(2) × (aP)−n] in the in vitro and in vivo trials is shown in Figure 3. The distribution in embryo cell number on Day 7 and the embryo length on Day 15 from the in vitro ET data (Shorten et al., 2018) are shown in Supplemental Figures S4 and S5 (https:// doi .org/ 10 .3168/ jds .2017 -14306). Mean Day 7 progesterone in cows with a recovered embryo on Day 15 in the in vitro trial was 5.11 ± 0.33 ng∙mL−1, whereas mean Day 7 progesterone in cows with a recovered embryo on Day 15 in the in vivo trial was 6.08 ± 0.22 ng∙mL−1 (P = 0.02) in yr 1 and 6.99 ± 0.22 ng∙mL−1 (P < 0.01) in yr 2. Day 7 progesterone was also significantly different between yr 1 and 2 (P = 0.01) and demonstrates the level of between trial and between year variability. However, the progesterone dependent doubling time relationship is not significantly different between 2014 and 2015 in vivo trials [the progesterone dependent doubling time is slightly greater in the 2015 trial than the 2014 trial, but this is because the estimate for the initial rate of embryo growth before d 5 (h1) was greater for the 2015 trial than the 2014 trial]. Although the variance in cell number on Day 7 is smaller in the in vitro study, the variance in embryo length on Day 15 in greater in the in vitro study (compare with Supplemental Figures S4 and S5; https:// doi .org/ 10 .3168/ jds .2017 -14306). This suggests that in vitro embryos are more sensitive to progesterone and potentially susceptible to greater stochastic perturbations in embryo growth.