In this paper we will examine the persistence of “misconceptions.” We used data from a longitudinal study of personal ideas in 24 students' thinking about ecological processes. The results show students often speaking about personal experiences dating from an early age, to which they had also referred in similar interviews conducted years before. These data are compared with results from a different study of middle school physics students' thinking about energy and steam engines. After the new learning had been “successfull” completed and assessed, old ideas returned. These findings are used to set up a theoretical basis for understanding the longitudinal results. Findings from memory studies are shown to explicate the long-term effects of the passage of time and prompts for the recall of scientific concepts.