David, Thanks. No, I was not trying to do time-series correlation of any kind. It is very simple. Say, you get three persons: A, B, C; and each of them consumes drinking some water last year -- for A, he drinks {a1,...,a12} gallon per month; for B, he drinks {b1,..., b12} gallon per month; and for C, he drinks {c1, ...., c12} gallon per month. I really want to see the pariwise correlation between monthly water consumptions bewtween A and B, between B and C, as well as between C and A... I was hoping there is a neat way in SAS without using proc transpose. I hope this explains my question better. Thank you, Paul