Past investment in offspring may be important in determining a parent's ability to reproduce in the future and, hence, should affect the relative value of current offspring. However, there have been surprisingly few clear tests of whether animals actually adjust parental care in response to diminished opportunities for future reproduction. We modified the experimental protocol of Sargent and Gross [Behav Ecol Sociobiol (1985) 17:43–45] to examine offspring desertion by mallards (Anas platyrhynchos), and decoupled the influence of past investment from expected current benefits by controlling for the effect of offspring age on clutch value. Using 9 years of nest mortality data, we accounted for the increasing prospects of egg survival with clutch age by calculating clutch sizes throughout incubation with equivalent expected benefits. Applying this approach, we experimentally reduced 203 clutches at two different incubation stages such that they had equivalent expected benefits but differed in the amount of past investment. Nest desertion rates did not differ between early- and late-incubated clutches that had equivalent expected benefits. Rather, the probability of desertion increased with the severity of the clutch reduction treatment. These results suggest that female mallards adjust parental care according to the expected benefits of current offspring, rather than to diminished prospects for future reproduction due to past investment. We further examined whether females assessed expected benefits on the basis of clutch size alone or clutch size adjusted for the age of the clutch. Using Akaike's Information Criterion, the most parsimonious model to explain the probability of deserting an experimentally reduced clutch included both the proportion of the clutch remaining and clutch age. Thus, female mallards appear to fine-tune their level of parental care not only according to the relative number of offspring in the clutch, but also to the increased prospects for offspring survival as they age.