https://communities.sas.com/t5/Statistical-Procedures/multiple-comparison-with-control/m-p/67937#M3287 I suggest you use the simulate option, because this is based on a multivariate t vector. In its simplest form, just use adjust=simulate; Sun, 22 May 2011 16:57:22 GMT lvm 2011-05-22T16:57:22Z https://communities.sas.com/t5/Statistical-Procedures/multiple-comparison-with-control/m-p/67936#M3286 I'm doing a multiple comparison with a control;<BR /> PROC GLM DATA=&lt;&gt; ;<BR /> CLASS &lt;&gt;;<BR /> MODEL &lt;&gt; =&lt;&gt;;<BR /> <BR /> LSMEANS &lt;&gt;/ PDIFF=CONTROL() ADJUST=DUNNETT CL;<BR /> LSMEANS &lt;&gt;/ PDIFF =CONTROL() ADJUST=BON CL;<BR /> LSMEANS &lt;&gt;/ PDIFF=CONTROL() ADJUST=SIDAK CL<BR /> LSMEANS &lt;&gt; /PDIFF=CONTROL () ADJUST =Multivariate t; ?<BR /> What option can I add for multivariate t ?<BR /> See table 39.7 of SAS/STAT(R) 9.2 user's guide second edition from documentation for similar thing. Message was edited by: Saszealot Fri, 20 May 2011 20:22:01 GMT https://communities.sas.com/t5/Statistical-Procedures/multiple-comparison-with-control/m-p/67936#M3286 Saszealot 2011-05-20T20:22:01Z https://communities.sas.com/t5/Statistical-Procedures/multiple-comparison-with-control/m-p/67937#M3287 I suggest you use the simulate option, because this is based on a multivariate t vector. In its simplest form, just use adjust=simulate; Sun, 22 May 2011 16:57:22 GMT https://communities.sas.com/t5/Statistical-Procedures/multiple-comparison-with-control/m-p/67937#M3287 lvm 2011-05-22T16:57:22Z