Summary: \(S\)-sample smooth goodness of fit tests are introduced. Testing may be hierarchical; we test if the first two samples are consistent with each other, and then if the \(s\) th, \(s=3,\dots, S\), sample is consistent with its predecessors. If all samples are consistent with having come from the same population, a convenient one-sample test for that distribution is available. If the samples are not consistent, a least significant difference (LSD) analysis can be performed to identify where the differences occur. The procedure provides a wealth of information.