AbstractWe test and quantify the (in)stability of farmer risk preferences, accounting for both the instability across elicitation methods and instability over time. We use repeated measurements (N = 1530) with Swiss fruit and grapevine producers over 3 years, using different risk preference elicitation methods (domain‐specific self‐assessment and incentivised lotteries). We find that farmers' risk preferences change considerably when measured using different methods. For example, self‐reported risk preference and findings from a Holt and Laury lottery correlate only weakly (correlation coefficients range from 0.06 to 0.23). Moreover, we also find that risk preferences vary considerably over time, that is, applying the same elicitation method to the same farmer in a different point in time results in different risk preference estimates. Our results show self‐reported risk preferences are moderately correlated (correlation coefficients range from 0.42 to 0.55) from one year to another. Finally, we find experiencing crop damages due to climate extremes and pests is associated with farmers becoming more risk tolerant over time in specific domains.