Identifying changes in the rate of background earthquakes is a common method to detect transients in seismic-activity rates, and these transients can often be linked to changes in physical properties of the crust or to the recording homogeneity of seismic networks. In this article, we consider the effect that various sources of uncertainty have on seismicity-rate estimation, and we describe a Monte Carlo method to compute the statistical significance of seismicity-rate changes. We consider real earthquake catalogs and synthetic catalogs that represent Poissonian background seismicity. We find that the choice of declustering algorithm and the variation in the related parameter values have the greatest impact on seismicity-rate-change calculations. This sensitivity is a direct consequence of the fact that there is no unique or universally applied approach to separate aftershocks and foreshocks from background events. Hypocenter location uncertainty also influences the statistical significance of seismicity-rate changes; its relevance depends on the ratio of the location uncertainty to the extension of the seismicity-rate anomaly. We find that magnitude uncertainty and edge effects associated with seismicity declustering have less impact than the other factors, but their influence is not negligible. In an example application, we suggest that, when uncertainties are considered, the reported seismicity-rate changes prior to the 1992 Landers and Big Bear mainshocks decrease significantly.