Dates setup to include mid-night to mid-night so in actuality the "stop" date count is not actually included in the sum.   Because of the above statement, I give a slightly modified output   proc format; value datef '03JUL2018'd-'04JUL2018'd = '03JUL2018-04JUL2018' '05JUL2018'd-'09JUL2018'd = '05JUL2018-09JUL2018'; run; proc summary data=have; class day; format day datef.; var count_of_surgeries; output out=_count_ sum=; run; If you have a longer table 1, you can turn this into a format without typing all the different dates via creating formats from the data set directly using the CNTLIN= option in PROC FORMAT. Example: https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/proc/p0owa4ftikc2ekn1q0rmpulg86cx.htm   Lastly ... advice ... 7/3/2018 could be July 3 or March 7, don't use this format, use 03JUL2018 which is unambiguous. ... View more

Hi @JenniferColeDH,   In some cases, including the example you've shown, KEEP= and DROP= dataset options (where name prefix lists are allowed) in conjunction with the asterisk notation can overcome that limitation of (PROC) SQL: PROC SQL; CREATE TABLE test(drop=uid) AS SELECT T1.NAME, T1.DATE, T2.* FROM LIST1 AS T1 INNER JOIN LIST2(keep=uid result:) AS T2 ON T1.UID = T2.UID; QUIT; ... View more