The aim of this study was to estimate the heritability (h2) of metabolite-based resilience biomarkers in dairy goat. Metabolites were measured repeatedly during two periods of stress: around parturition and during a feeding challenge (48 h with straw only) in early first lactation in two INRAE facilities (P3R Bourges and Paris). The 4 blood metabolites were: glucose (Glu), beta-hydroxy-butyrate (BOHB), urea and non-esterified fatty acids. The 14 milk metabolites were: BOHB, Glu, urea, glucose-6-phosphate (Glu6P), galactose, isocitrate, glutamate, NH2-groups, lactate dehydrogenase (LDH), choline, malate, urate, triacylglycerol, and cholesterol. The metabolite trajectories were described both by simple mean concentration per challenge period and by a functional PCA method. Variance components were estimated using an animal model (wombat®) for blood metabolites around kidding, and for trajectories of blood and milk metabolites upon the feeding challenge on 201, 228 and 138 goats respectively. The model included the fixed effects of facility and year. The total pedigree included 1,148 animals. We found 25 blood and milk metabolites parameters (out of 159) that were significantly heritable (h2-2SE >0) with h2 estimates ranging from 0.30 (±0.03) to 1.00(±0.25) with SE from 0.03 to 0.33. Among them: milk urate, BOHB, LDH, and Glu6P during feeding challenge and blood Glu around kidding. The heritability estimation of 250-d milk yield, fat and protein content were respectively 0.26 (±0.16), 0.53 (±0.19) and 0.68 (±0.19), showing the consistency of our dataset despite the small number of individuals. These results show the potential of metabolite-based biomarkers for genetic selection of resilience. This study has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No 772787 (SMARTER