Hello Community,   I am working with proc compare and would like to output records that are in the base dataset and not found in the compare dataset and the same if records are found in the compare dataset and not in the base dataset. In my example have1 dataset is the master dataset never to be updated, if something in have2 is different then an issue accrued during data entry,  The datasets from the proc compare output H1_H2_diff_id and H2_H1_diff_id give me what I would expect for records needing attention in have2 that differ from have1.     How can I make proc compare output records found in one dataset and not the other dataset.       data have1; input id $ date mmddyy10.; cards; 12 11/1/2009 1456 1/1/2008 1 12/12/2012 2158974 4/4/2018 2158984 5/4/2018 2458974 9/4/2018 2158972 4/1/2018 2587911 10/1/2018 2587915 10/8/2018 ; proc sort data=have1; by id; run; data have2; input id $ date mmddyy10.; cards; 1456 1/1/2008 1 12/12/2012 2158974 4/4/2018 2158984 5/4/2018 2587915 10/7/2018 2458974 9/4/2018 2158972 4/1/2018 987588121 10/31/2018 ; proc sort data=have2; by id; run; /* With the by ID I get what I want when records have different values. but I would like to output records found in base that are not found in compare to dataset inbase_not_compare and if records are in compare that are not in base to another dataset incompare_not_base. Can this be done in the proc compare process? */ proc compare base=have1 compare=have2 out=h1_h2_diff_id outall outnoequal; by id; run; proc compare base=have2 compare=have1 out=h2_h1_diff_id outall outnoequal; by id; run; proc compare base=have1 compare=have2 out=h1_h2_diff outall outnoequal; run; proc compare base=have2 compare=have1 out=h2_h1_diff outall outnoequal; run; Thank you for you assistance.         ... View more