Deleterious alleles may be removed (purged) by natural selection in populations undergoing inbreeding. However, there is controversy regarding the effectiveness of purging in reducing the extinction risk due to inbreeding, particularly in conservation contexts. We evaluated the effects of purging on the extinction risk due to inbreeding in Drosophila melanogaster using two base populations, an outbred population (non-purged) and four-way crosses between highly inbred lines derived from the same population (purged). The inbred lines used in the four-way crosses were previously subjected to 20 generations of full-sib mating. The impact of full-sib inbreeding over a further 12 generations was compared in 200 populations from each of the two base populations. There was a small and non-significant difference between the extinction rates at an inbreeding coefficient of 0.93 in the non-purged (0.74 ± 0.03) and purged (0.69 ± 0.03) treatments. This is consistent with other evidence indicating that the effects of purging are often small. Purging using rapid inbreeding in very small populations cannot be relied upon to eliminate the deleterious effects of inbreeding.