The ability to record transcriptional events within a cell over time would help to elucidate how molecular events give rise to complex cellular behaviours and states. However, current molecular recording technologies capture only a small set of defined stimuli. Here we use CRISPR spacer acquisition to capture and convert intracellular RNAs into DNA, enabling DNA-based storage of transcriptional information. In Escherichia coli, we show that defined stimuli, such as an RNA virus or arbitrary sequences, as well as complex stimuli, such as oxidative stress, result in quantifiable transcriptional records that are stored within a population of cells. We demonstrate that the transcriptional records enable us to classify and describe complex cellular behaviours and to identify the precise genes that orchestrate differential cellular responses. In the future, CRISPR spacer acquisition-mediated recording of RNA followed by deep sequencing (Record-seq) could be used to reconstruct transcriptional histories that describe complex cell behaviours or pathological states.