Abstract Study question Can non-invasive metabolic imaging detect variations in cumulus cell metabolic parameters associated with a viable pregnancy of the corresponding embryo? Summary answer Noninvasive metabolic imaging can detect differences in the cumulus cell metabolic signatures between embryos that led to a viable pregnancy and those that did not. What is known already Bidirectional metabolic cooperativity between the human oocyte and its surrounding cumulus cells is essential for the oocyte to acquire full developmental competency. However, the relationship between cumulus cell metabolism and oocyte viability is not well established. Metabolic imaging uses Fluorescence Lifetime Imaging Microscopy (FLIM) to non-invasively measure autofluorescence of the endogenous molecules, NADH and FAD+, which are essential coenzymes for cellular respiration and glycolysis. This technique enables quantitative information for these coenzyme concentrations and regarding metabolite enzyme engagement. We have previously shown that this technique is an effective tool for quantitatively measuring metabolic state of mouse embryos. Study design, size, duration Cumulus cell clusters (n = 617 from 193 patients) were dissected from cumulus-oocyte complexes prior to insemination or ICSI, vitrified, warmed and their metabolic function assessed. We conducted a prospective observational study to evaluate to what extent cumulus cells from an oocyte that led to a viable pregnancy (presence of a viable fetus >7 weeks gestation) after transfer of the corresponding embryo metabolically differed from those that did not. We also evaluated the associations with embryo morphology. Participants/materials, setting, methods Cumulus cell metabolism was assessed non-invasively using FLIM to measure the autofluorescence of NADH and FAD+. Overall a single FLIM measurement provides a total of 8 metabolic parameters (4 for NADH and 4 for FAD+). An additional parameter, the Redox Ratio was also acquired (NADH intensity / FAD+ intensity). We used multilevel models to investigate the association of cumulus cell metabolic parameters with the morphology of the corresponding embryo and clinical outcome. Main results and the role of chance Of the cumulus samples analyzed, 75 corresponded to embryos that did not result in a viable pregnancy, and 24 that did so. Significant associations were observed between cumulus cell FAD+ fraction bound to enzyme (p = 0.007), FAD+ long lifetime (p = 0.01) and FAD+ short lifetime (p < 0.001) and the clinical outcome of the corresponding embryo. These significant associations held up after controlling for age. We used a support vector machine algorithm to distinguish between embryos that led to a viable pregnancy and those that did not. The optimum hyperplane derived from a support vector machine algorithm predicted whether a sample with random cumulus cell metabolic parameters will lead to a viable pregnancy or not with an accuracy of 80%. Embryo morphological assessments were stratified as excellent, good, fair and poor. We found no significant associations between cumulus cell metabolic signatures and embryo morphology evaluated on day 3. Significant associations of FAD+ short lifetime (p < 0.001) and day 5 embryo morphology were found. However, these associations were not significant after controlling for age. Limitations, reasons for caution Although we observed significant variations in metabolic parameters, further studies with larger sample sizes are required. Despite our validation studies showing no significant effect of vitrification on cumulus cell metabolic parameters, analyses with fresh clusters are needed to confirm our results. Wider implications of the findings Noninvasive FLIM imaging detects metabolic variations of cumulus masses and their association with embryo viability. The ability to correlate metabolic measurements of cumulus clusters, in combination with embryo morphology assessments and patient clinical characteristics, with embryo fate paves the way for this approach to be used in a clinical setting. Trial registration number 5RO1HD092559-03