The Irish hare Lepus timidus hibernicus is of priority conservation concern in Northern Ireland, based on a perceived decline in abundance during the 20th century and on two surveys in 1994–1996 and 2002 that suggested low densities. We reanalysed a 10-year time series [Northern Ireland Rabbit Survey (NIRS)] that has been key evidence of decline, to derive an annual index of abundance during 1986–1995. This showed a rise to a double peak followed by a trough. We compared these Northern Ireland data with two datasets for Lepus timidus scoticus. A 40-year annual index derived from hare bags on Scottish shooting estates [National Game-Bag Census (NGC)] showed strong fluctuations, each lasting several years. Autocorrelation analysis of the NGC data showed elements of cyclical behaviour, with distinct phases and a 16-year period. A previously published 30-year survey of introduced L. t. scoticus in northern England also showed large multi-year fluctuations and phasic behaviour, albeit on a shorter time scale. The NIRS index was strongly correlated with NGC data for the same years, suggesting that it too reflected part of a natural population dynamic. The Species Action Plan for L. t. hibernicus in Northern Ireland includes the aims of maintaining the existing range and demonstrating increase by 2005, and of doubling the population by 2010. We suggest that these aims and the supporting monitoring strategy are inappropriate for a species that shows multi-annual fluctuations in abundance throughout its geographical range. It is important to distinguish self-correcting population processes from trends requiring remedial conservation action.