One of the most striking developments in understanding science learning has been the discovery of the extent and persistence of the naive conceptions about the natural world students bring with them to the classroom. In physics and other sciences, students (even those who do well on textbook problems) often do not apply principles they have learned to predicting and describing actual physical events. Investigations have revealed that these students' failures were not due to an absence of theories, but to the persistence of naive theories brought with them to science classes, theories that stand in marked contrast to what they are expected to learn. Evidence is accumulating that these naive theories and the distortions they engender in students' comprehension of instruction are among the principal causes of their failure to achieve understanding in science. Discussed in this paper are: (1) the characteristics of naive conceptions; (2) the influence of naive conceptions on students' interpretations of instructional events; and (3) the implications of this research for designed instruction to facilitate the reconciliation of naive conceptions with scientific theories. (Author/JN)