Cosmogenic exposure dating has sometimes been used to identify moraines associated with short-lived climatic events, such as the Younger Dryas (12.9-11.7 ka). Here we point out two remaining challenges in using exposure dating to identify moraines produced by abrupt climate changes. Specifically, (1) a commonly applied sampling criterion likely yields incorrect exposure dates at some sites, and (2) geomorphic processes may introduce bias into presently accepted nuclide production rate estimates. We tit a geomorphic process model that treats both moraine degradation and boulder erosion to collections of exposure dates from two moraines that were deposited within a few thousand years of the Younger Dryas. Subsampling of the modeled distributions shows that choosing boulders for exposure dating based on surface freshness yields exposure dates that underestimate the true age of the moraine by up to several thousand years. This conclusion applies only where boulders do not erode while buried but do erode after exhumation. Moreover, one of our fitted data sets is part of the global nuclide production rate database. Our fit of the moraine degradation model to this data set suggests that nuclide production rates at that site are several percent higher than previously thought. Potential errors associated with sampling strategies and production rate estimates are large enough to interfere with exposure dating of moraines, especially when the moraines are associated with abrupt climate changes. We suggest sampling strategies that may help minimize these problems, including a guide for determining the minimum number of samples that must be collected to answer particular paleoclimate questions.