This paper outlines and evaluates two techniques for testing hypotheses about mean elasticities obtained from linear regression equations. These tests are often necessary so that the implications of the estimates can be fully appreciated. For example, we are often interested in whether a given elasticity is greater or smaller than unity.