Cancer immunotherapy with checkpoint blockade has become standard of care treatment for numerous cancer types. Despite this, robust predictive biomarkers are lacking. There is increasing evidence that the host microbiome is a predictor of immunotherapy response, although the optimal host microbiome has not been defined. Metabolomics is a new area of medicine that aims to analyze the metabolic profile of a biological system. The microbiome-derived metabolome (fecal and serum) represents the end products of microbial metabolism and these may be functionally more important than the distinct bacterial species that comprise a favorable microbiome. Short-chain fatty acids (SCFA) are metabolites produced by gut microbiota and have a role in T cell homeostasis, including differentiation of regulatory T cells. Recent studies have confirmed differential expression of SCFA for immunotherapy responders compared with non-responders. We propose that the microbiome metabolome, with a focus on SCFA may be a novel predictive biomarker for immunotherapy efficacy.