Automated and unbiased methods of non-invasive cell monitoring able to deal with complex biological heterogeneity are fundamentally important for biology and medicine. Label-free cell imaging provides information about endogenous fluorescent metabolites, enzymes and cofactors in cells. However extracting high content information from imaging of native fluorescence has been hitherto impossible. Here, we quantitatively characterise cell populations in different tissue types, live or fixed, by using novel image processing and a simple multispectral upgrade of a wide-field fluorescence microscope. Multispectral intrinsic fluorescence imaging was applied to patient olfactory neurosphere-derived cells, cell model of a human metabolic disease MELAS (mitochondrial myopathy, encephalomyopathy, lactic acidosis, stroke-like syndrome). By using an endogenous source of contrast, subtle metabolic variations have been detected between living cells in their full morphological context which made it possible to distinguish healthy from diseased cells before and after therapy. Cellular maps of native fluorophores, flavins, bound and free NADH and retinoids unveiled subtle metabolic signatures and helped uncover significant cell subpopulations, in particular a subpopulation with compromised mitochondrial function. The versatility of our method is further illustrated by detecting genetic mutations in cancer, non-invasive monitoring of CD90 expression, label-free tracking of stem cell differentiation, identifying stem cell subpopulations with varying functional characteristics, tissue diagnostics in diabetes, and assessing the condition of preimplantation embryos. Our optimal discrimination approach enables statistical hypothesis testing and intuitive visualisations where previously undetectable differences become clearly apparent.