Metastasis represents the deadliest outcome in cancer, leading to the vast majority of cancer-related deaths. Understanding the progression from micro-to macro-metastasis might improve future therapeutic strategies aimed at blocking metastatic disease. However, the difficulty of investigating vital, clinically undetectable, micro-metastases hindered our capacity to unravel phenotypic determinants of micro-metastases. In this work, we leveraged indocyanine green (ICG) dye to detect small sized liver micro-metastases across several cancer models. We exploited a method for infrared fluorescence scanning of fresh tissue and coring of cancer micro-metastases and succeeded in processing them for single-cell RNA sequencing. Our analysis revealed that distinct liver micro-metastases upregulate both shared and specific genes that can successfully predict breast cancer patient prognosis. Moreover, the ontology classification of these genes allowed the validation of several pathways, namely interferon response, extracellular matrix remodeling, and antioxidant response in metastatic progression. Ultimately, we showed that ICG can be successfully used to quantify breast cancer micro- and macro-metastases to lungs, which we showed to be abrogated through inhibition of H 2 O 2 -producing enzyme monoamine oxidase. Therefore, the ICG approach allowed us to identify not only determinant of breast cancer metastatization, but also to assess the therapeutic efficacy of targeting these genes which can be further investigated in clinic.