A consideration of the theoretical relationship between vegetational percentages and the percentages of pollen in sediments makes it obvious that where species differ in the amounts of pollen they contribute to sediments, pollen diagrams cannot be interpreted quantitatively by means of traditional methods. Under ideal conditions it should be possible, however, to interpret fossil pollen percentages objectively by means of correction for observed differences in pollen contribution. This method is described and demonstrated by its application to the Pine Pollen Zone in a pollen diagram from Vermont. The results imply that our previous subjective vegetational and climatic interpretation of this zone has been without validity. The departures of the real situation from the conditions of an ideal model are used to illustrate a general discussion of some of the subtle but potentially important sources of error that must be evaluated before pollen analysis can be used as a reliable paleoecologic technique.