Purpose Financial toxicity is a contributor to the psychosocial burden of cancer care. There is no consensus measure of financial toxicity, however, recent studies have adopted the COmprehensive Score for financial Toxicity (COST). Despite its growing popularity, data on the responsiveness to change of the COST are lacking. To address this gap in the literature, we performed a sequential mixed-methods study of people with multiple myeloma.Methods In the quantitative phase of the study, we collected COST scores at two time points approximately 8 weeks apart from 72 patients. In the qualitative phase, we conducted semi-structured interviews with a subset of 12 patients who reported the largest changes in COST scores. The qualitative data were analyzed using a deductive coding scheme developed using the Framework Method in the context of a commonly cited conceptual model of financial toxicity.Results The median absolute change in COST scores was 4 points (IQR 2–6). Only 13% of the sample had the same COST scores at both assessments; 38% had an improved score and 50% had a worsened score. Only, seven of the twelve patients (58%) interviewed reported changes to one or more of the constructs in the conceptual model of financial toxicity. Most commonly, changes to out-of-pocket costs were reported (5/12). Changes to non-medical expenses (n = 2) and subjective financial distress without changes to objective financial burden (n = 2) were also reported.Conclusion Additional research is needed to explicate changes in COST scores over time.