An integrated overview is provided of selected multiple comparison methods, with an emphasis on the sequentially rejective Bonferroni (SRB) test and its modifications. Multiple comparisons of means is a frequently used, and yet confusing, method in educational and psychological research. Seven multiple comparison methods are reviewed: (1) the classical Bonferroni t test (CB); (2) Tukey's HSD test (1953); (3) H. Scheffe's test (1953); (4) C. W. Dunnett's test (1955); (5) the SRB; (6) a modification of the SRB (MSRB1); and (7) another modification of the SRB (MSRB2). It is concluded that the SRB introduced by S. Holm (1979) is flexible and easy to use in various research situations and has more power than does the CB procedure. The modified SRB procedures described give more power than does the simple SRB, although they have restricted applications. If a researcher prefers a simple procedure with reasonable power in conducting all pairwise comparisons, either the Tukey test or the SRB seems to be the best choice. All of the described procedures control experimentwise error rates at the nominal level of significance. The choice among these procedures depends on the experimenter's decisions about the type of contrast that is the primary interest, the most powerful procedure, and the desired computational simplicity. Some of the recent multiple comparison procedures are presented geometrically to aid visual understanding of these complex methods. Eight tables and seven figures are included. (SLD)