This sounds so much like a classroom exercise I'm going ask if it is.   The main reason I want to know if it is a school exercise is, having worked with BRFSS data for approaching 20 years, the complex weighting schemes generally mean that procedures that do not account for the sample design generally are not suitable for analysis.   As a minimum you should account for the weights associated. That is what the Weight statement is for.   Easiest is to have an ODS OUTPUT statement for each table that you want to output, if that's the assignment. This line: TABLES race adult sex * Asthma / NOROW NOCOL NOFREQ CHISQ CL;   possibly should be TABLES (race adult sex) * Asthma / NOROW NOCOL NOFREQ CHISQ CL; if you want to cross the first 3 variables with the asthma variable. Since you have arequirement to include a row percent you might want to consider dropping that NOROW. That generates 3 tables of output. (HINT)   ... View more