"Diagnostic delay," the duration of symptoms or the symptom to diagnosis interval (SDI), are highly complex variables that reflect the behavior of the patient and the attending physician, tumor biology and host-tumor interactions, the functioning of the health care system, and sociocultural norms. In addition to tumor stage, other variables mediate the relationship between duration of symptoms and survival; clinical and epidemiologic procedures to measure them must be improved. Largely at odds with clinical and common wisdom, decades of research have shown that often SDI is not associated with tumor stage and/or with survival from cancer. It would be relevant to increase evidence in support of the notion that, for each type of tumor, there is a positive relationship between the length of the presymptomatic and the symptomatic phases. SDI could then be used to classify tumors according to their likelihood of being detected early when still asymptomatic. Also, tumors could be classified according to the ratio of the median SDI to the median survival (SDI to survival ratio, SSR), which may estimate the relative likelihood for clinical lead-time bias. If adhering to rigorous methodologic standards, proteomic analyses of early-stage cancers might provide new insights into changes that occur in early phases of tumorigenesis. More real examples are needed of uses of pathologic and genomic data to study mechanisms through which SDI influences-or fails to influence-prognosis. The degree of correlation between proteomic patterns and classic semiology constitutes an area of interest in itself; their respective correlations with cancer prognosis should be assessed in properly designed epidemiologic studies.