If your PROC MIXED code contains a RANDOM statement and both of the following conditions are met:

1. there are missing values in any of the CLASS variables or covariates used on the CLASS, MODEL, or RANDOM statements
2. the RANDOM statement uses a GROUP= option with no SUBJECT= option on that statement

then the output from PROC MIXED might be incorrect. The above two conditions must hold for the output to be incorrect, but the two conditions above do not guarantee that the output will be wrong.

To circumvent the problem, either drop the observations with missing values in the predictor variables before running PROC MIXED or (if possible) rewrite your RANDOM statement so that it contains a SUBJECT= effect.

Missing values in the dependent variable on the MODEL statement do not trigger this problem. There must be missing values in the predictor variables for the problem to occur.

Select the Hot Fix tab in this note to access the hot fix for this issue.

Operating System and Release Information

A fix for SAS 9.1.3 (9.1 TS1M3) for this issue is available at:

A fix for SAS 9.1.3 (9.1 TS1M3) with Asian Language Support (DBCS) for this issue is available at :