Perhaps this narrative regarding DBMS might help: DBMS=identifier specifies the type of data to import. You can import JMP files (DBMS=JMP) or delimited files. To import a tab-delimited file, specify TAB as the identifier. To import any other delimited file that does not end in .CSV, specify DLM as the identifier. For a comma-separated file with a .CSV extension, DBMS= is optional. The IMPORT procedure recognizes .CSV as an extension for a comma-separated file. See:DBMS Identifiers Supported in Base SAS for more information about identifiers for this option. ... View more

Could it be that SAS does not recognize the .DTA file extension? The following example appears online. Maybe it can help you, specifically the DBMS and DELIMITER options. I am sorry but I have only imported CSV and EXCEL files into SAS. proc import datafile='c:\temp\tab.txt'   out=class   dbms=dlm   replace;   delimiter='09'x;   datarow=5; run; ... View more

The simplest answer is a combination of 2 answers: data have; input Subject $ Value1    value2 value3  value4; datalines; 10001     3        4       5       8 10007     4        6       4       9 10009     4        9      10       3 run; data need;    set have;    value = max(of value1 - value4); run; the dataset NEED will look as follows: 10001     3        4       5       8     8 10007     4        6       4       9     9 10009     4        9      10       3     10 Message was edited by: Amin Bardai ... View more