The non-associative framework on fear argues that evolutionary-relevant fears, such as heights and water, are largely innate and do not require associative conditioning experiences to develop (Menzies & Clarke, 1993a, b; Poulton & Menzies, 2002a). However, this framework has been criticized for its reliance on retrospective recall for empirical support (Mineka & Öhman, 2002), which has been found to be highly unreliable (Taylor, Deane, & Podd, 1999). Thirty height-fearful undergraduate students completed the Origins Questionnaire-II (OQ-II; Menzies & Parker, 2001) and were classified into one of the several associative and non-associative pathways of fear onset based on their responses. A control group of 43 non-fearful students completed a modified version of the OQ-II to report any past experiences with heights. To examine the stability of these responses over time, the same questionnaires for both groups were completed again three months (Time 2) and 12 months (Time 3) after the initial administration of the test, along with measures of fear severity. Results showed that neither associative nor non-associative accounts took precedence over the other in explaining the onset of height fear. Instabilities in pathway ascriptions were observed in 18.18% of cases over three months (between Time 1 and Time 2), and 27.27% of cases over nine months (between Time 2 and Time 3). The theoretical and practical implications of the results are discussed with consideration of some of the study’s procedural and instrumental limitations. In light of these limitations, this study identified a substantial role of non-associative pathways on the development of height fear, and provided further support for the limitations of retrospective recall for ascertaining the pathway to fear onset.