Likelihood ratio considerations are used to derive uniform bounds on the acceptance and rejection regions of the family of locally most powerful permutation tests against Type 1 censored alternatives. These bounds are applicable to radio immune assays and to survival and reliability studies.

With most samples, the proportion of rearrangements that lie in the indifference region between the bounds is small enough to permit effective decision making. As the sample size increases, the associated permutation test converges in probability to a UMP unbiased test. But these also exist samples for which the indifference region is many times larger than the rejection region.