Increased understanding of the phenotype and genotype abnormalities associated with cancer including cytogenetic, immune and molecular features has facilitated development of tests to identify and quantify residual cancer cells not detected by conventional methods. Progress is especially rapid in haematological compared with solid cancers. Starting from modestly accurate assays for measurable residual disease (MRD), testing is being used at many centres and available from commercial laboratories. The hope is results of these tests will provide clinically meaningful predictive and/or prognostic data and inform therapy decision-making. We created a recent online collection of articles on MRD focusing on critical analyses of the extent to which these goals have been achieved in haematological and solid cancers – see www.nature.com/collections/mrd.

Leukemia has always been interested in MRD. The change from the subjective term minimal residual disease to measurable residual disease, an objective term when assay conditions are specified, derives from an article published in Leukemia 10 years ago by Goldman and Gale [1] and is now widely-adopted.

Efficacy and use of results of MRD-testing varies between different blood cancers. The best established is in chronic myeloid leukaemia (CML) where a canonical mutation, BCR::ABL1, facilitated development and standardisation of MRD assays and where results of MRD-testing are routinely used to direct and when to stop therapy. Three Perspectives in this Collection discuss clinical use of MRD-testing in haematological cancers. Ofran & Rowe review use of MRD-testing in B-cell acute lymphoblastic leukaemia (ALL) discussing its value compared with cytogenetics testing [2]. Next, Del Giudice & Foà summarise where we stand with MRD-testing in chronic lymphocytic leukaemia (CLL) [3]. In a 3rd Walter briefly reviews the current role of MRD-testing in acute myeloid leukaemia (AML) [4]. The last Perspective by Chen and colleagues considers MRD-testing in haematological and solid cancers [5]. The authors discuss limitations inherent in MRD-testing and interrogate the strength of data supporting the use of MRD data for therapy decisions [5].

Our hope is readers of the Collection will benefit from the nuanced insights and practical implications provided by these articles, especially the critical analyses. Readers interested in more in-depth discussions of MRD-testing in haematological cancers may want to consult these excellent references in Leukemia [2, 6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25]. As Chen and colleagues noted quoting Lord Kelvin: When you can measure what you are speaking about and express it in numbers, you know something about it.