RMNP2 is an easy-to-use FORTRAN program for the analysis of repeated measures using the non-parametric two-sample tests of Wei and Lachin (J. Am. Stat. Assoc. 79 (1984) 653-661) and Wei and Johnson (Biometrika 72 (1985) 359-364). The program compares two groups of subjects or experimental units when measurements are obtained at multiple time points, or under multiple conditions, from each subject. A strength of the methodology is that subjects with missing responses at one or more time points can be included in the analysis, under the assumption that the missing value mechanism is independent of the response. In contrast to other methods that require parametric assumptions concerning the distribution of the outcome variable, RMNP2 is applicable when the response variable is continuous but not normally distributed. The program is also useful in the analysis of ordered categorical outcomes when the number of possible responses is too large to permit application of general categorical data methodology. The program can be run on microcomputers, workstations and mainframe computers. Two examples illustrating the use and features of RMNP2 are provided.