Bill measurement is important in many feeding studies. Traditionally, two alternative bill measures have been proposed: bill length measured from the tip to skull (i.e., total culmen), mainly used in Europe, and bill length measured from the tip to the anterior edge of nostrils, used both in Europe and America. However, the correlation between both measures as well as the analysis of what they are in fact measuring have not been explored yet. The aim of this paper is to test in the Citril Finch (Serinus citrinella) whether measurement of bill length from the tip to the anterior edge of nostrils and total culmen are measuring different components of bill length (e.g., ramphotheca vs. premaxilla), and to assess which is the best measure in relation to their measurement error. This is specially relevant in granivorous birds for which total culmen length differs from ramphotheca length. The correlation between bill length to nostrils and total culmen was moderate (r = 0.75). Multiple regression of total culmen (dependent variable) on the bill components included ramphotheca length (R² = 52%) and marginally premaxilla length to nostrils (R² change = 6%, P = 0.06). When using bill length to nostrils as dependent variable it also included ramphotheca length (R² = 60%) and premaxilla length to nostrils (R² change = 16%, P < 0.001). Measurement error was low for bill length to nostrils (ME = 1.6%) but high for total culmen (ME = 26.3%), making the latter measure not reliable. The replicability of total culmen is probably decreased in granivorous birds, compared to other passerines, by the fact that in these species the abrupt edge of the ramphotheca differs from the point at which the culmen meets the skull, which hinders the location of the end of culmen. All of this highly recommends that, specially in granivorous birds, the length of bill should be measured from nostril. Our results also show that both bill length measures are mainly measuring ramphotheca.