A dynamic peanut simulation model was used to quantify climatic risk to peanut production in Northern Australia. We demonstrate how district yield information can be usefully combined with simulation results to assess objectively impact and causes of climatic variability on production. Our analysis shows that the rapid expansion of the peanut industry in the region corresponded with relatively stable, above average yields caused by that period in the historical record having above-average and less variable summer rainfall. During this period the timing and amount of rainfall was such that yields higher than average could often be achieved, and harvests were only rarely interrupted by prolonged wet periods. These conditions created unrealistically high expectations of yields by producers, and when the climate was more variable during the 1980s, it was perceived as a greater deviation from the norm than justified by the long-term record.