This paper discusses the concepts of rate and risk in epidemiology, particularly in the context of competing risks. In all-cause mortality studies, rate and risk are closely related, but this one-to-one correspondence is lost in the presence of competing risks. The paper explains that in such cases, the naive Kaplan–Meier estimator, which treats competing events as censored, is biased. Additionally, the association between covariates and cause-specific hazards may differ from their association with cumulative incidence. The paper illustrates these concepts using data from a stem cell transplantation study, where relapse and non-relapse mortality are competing risks. It also discusses the implications of these findings for statistical modeling and inference in the presence of competing risks. The paper highlights the importance of distinguishing between cause-specific hazards and cumulative incidence, and emphasizes that models for cause-specific hazards may not directly translate to models for cumulative incidence. The paper also discusses the Fine-Gray model, which directly links cumulative incidence to explanatory variables, and notes that its parameters are not easily interpretable. The paper concludes that both rates and risks remain useful in the analysis of competing risks, with cause-specific hazards being more relevant when the disease etiology is of interest, while cumulative incidences are more relevant for prediction. The paper also addresses the concept of independent competing risks, which is difficult to verify but not necessary for inference. The paper is supported by funding from various organizations and acknowledges the European Group for Blood and Marrow Transplantation for providing the data.This paper discusses the concepts of rate and risk in epidemiology, particularly in the context of competing risks. In all-cause mortality studies, rate and risk are closely related, but this one-to-one correspondence is lost in the presence of competing risks. The paper explains that in such cases, the naive Kaplan–Meier estimator, which treats competing events as censored, is biased. Additionally, the association between covariates and cause-specific hazards may differ from their association with cumulative incidence. The paper illustrates these concepts using data from a stem cell transplantation study, where relapse and non-relapse mortality are competing risks. It also discusses the implications of these findings for statistical modeling and inference in the presence of competing risks. The paper highlights the importance of distinguishing between cause-specific hazards and cumulative incidence, and emphasizes that models for cause-specific hazards may not directly translate to models for cumulative incidence. The paper also discusses the Fine-Gray model, which directly links cumulative incidence to explanatory variables, and notes that its parameters are not easily interpretable. The paper concludes that both rates and risks remain useful in the analysis of competing risks, with cause-specific hazards being more relevant when the disease etiology is of interest, while cumulative incidences are more relevant for prediction. The paper also addresses the concept of independent competing risks, which is difficult to verify but not necessary for inference. The paper is supported by funding from various organizations and acknowledges the European Group for Blood and Marrow Transplantation for providing the data.