The ROC curve is a graphical representation of the trade-off between the sensitivity (true positive rate) and specificity (false positive rate) of a diagnostic test or model across different thresholds. The curve plots sensitivity on the y-axis and 1-specificity on the x-axis.