Mating preference can evolve as a side effect of ecological adaptation and simultaneously contribute to speciation in certain scenarios. However, theoretical predictions have been difficult to test experimentally because there has not been any simple way to relate empirical and theoretical parameters. Recently, it has been shown that the r coefficient and other statistics can be used to estimate mating preference in wild-captured mating pairs. In the present work we use these estimators to test whether mating preference has significantly increased in a context favorable for ecological divergence. We perform the study using two groups of samples from populations of the marine gastropod Littorina saxatilis: 1) a group of bimodal populations affected by ecological divergence, and 2) a group of unimodal populations. The results show that mating preference (unlike the trait affected by the preference) does not increase when two ecotypes meet and mate in sympatry, which suggests that mating preference must be controlled by strong stabilizing natural selection, and/or limited by mating cost. This would suggest that mating preference would rarely contribute to speciation under the face of gene flow in these ecotypes.