Many higher education studies have examined linear relationships between student experiences and outcomes, but this assumption may be questionable. In two notable examples previous research that assumed a linear relationship reached different substantive conclusions and implications than did research that explored non-linear associations among the same constructs. Indeed, many relationships between college experiences and outcomes may actually be curvilinear; this study explored that possibility within a large, multi-institutional, longitudinal dataset. As expected, most of the significant positive relationships were accompanied by significant curvilinear associations, such that the magnitude of the relationship decreased with higher levels of involvement.