1: \begin{abstract}

2: This paper provides conditions under which subsampling and the

3: bootstrap can be used to construct estimators of the quantiles of the

4: distribution of a root that behave well uniformly over a large class of

5: distributions $\mathbf P$. These results are then applied (i) to

6: construct confidence regions that behave well uniformly over $\mathbf

7: P$ in the sense that the coverage probability tends to at least the

8: nominal level uniformly over $\mathbf P$ and (ii) to construct tests

9: that behave well uniformly over $\mathbf P$ in the sense that the size

10: tends to no greater than the nominal level uniformly over $\mathbf P$.

11: Without these stronger notions of convergence, the asymptotic

12: approximations to the coverage probability or size may be poor, even in

13: very large samples. Specific applications include the multivariate

14: mean, testing moment inequalities, multiple testing, the empirical

15: process and $U$-statistics.

16: \end{abstract}