Thank you for your help! There is one more piece to the puzzle that I failed to mention. I am reading in multiple files like this, and they have varying numbers of columns.
I had used the 'datarow=' option for proc import, which resulted in var1, var2, etc. with the column headers in observation 1. Can I loop through a name assignment?
Could I do something like (pseudocode, not sure exactly how to do this)
for K = 1 to number of variables
vname(varK) = value(observation 1, column K)
next K
Is there some way of ignoring or deleting the first 20 observations, so that the 21st observation with the column headers now becomes the 1st observation, so that I can use the 'getnames=yes' option?
My ultimate solution would be to somehow
1) search the data for a unique known column header that will always appear in the file,
2) ignore prior rows/observations, and
3) read in the column headers as variable names with the data directly below this.
This way I would be able to deal with files containing variable "fluff" prior to my column headers, and variable number of columns.