Single-cell sequencing methods have enabled in-depth analysis of the diversity of cell types and cell states in a wide range of organisms. These tools focus predominantly on sequencing the genomes1, epigenomes2 and transcriptomes3 of single cells. However, despite recent progress in detecting proteins by mass spectrometry with single-cell resolution4, it remains a major challenge to measure translation in individual cells. Here, building on existing protocols5–7, we have substantially increased the sensitivity of these assays to enable ribosome profiling in single cells. Integrated with a machine learning approach, this technology achieves single-codon resolution. We validate this method by demonstrating that limitation for a particular amino acid causes ribosome pausing at a subset of the codons encoding the amino acid. Of note, this pausing is only observed in a sub-population of cells correlating to its cell cycle state. We further expand on this phenomenon in non-limiting conditions and detect pronounced GAA pausing during mitosis. Finally, we demonstrate the applicability of this technique to rare primary enteroendocrine cells. This technology provides a first step towards determining the contribution of the translational process to the remarkable diversity between seemingly identical cells.