In this paper we are concerned with comparing ordered treatment effects with a control in a randomized block design with multiple observations per cell. Two nonparametric procedures for detecting which treatment are better than the control are proposed and compared. An example is given and the results of a Monte Carlo power study are discussed.