I'm trying to identify the relationship between a series of X variables and a series of Y variables. If an X variable is correlated with the Y variable, is it then appropriate to do a regression to "get more from the data", or is this only the case if my hypothesis are about predictive relationships?

Soo confused!