Many studies of insect behavior in relation to chemicals from different plants require the use of statistical tests for determining whether attraction, repulsion, or neutrality is significant. In this paper, we propose a simple and efficient nonparametric statistical procedure for testing hypotheses about insect preference in two-choice experiments with individual insects. Three criteria are provided in order to detect the following alternatives to neutrality: (i) general attraction (or repulsion) in a sample population, (ii) strong variability of the individual responses, and (iii) slight but similar behavioral effects. Illustrations are given from two-choice experiments concerning the European corn borer (Ostrinia nubilalis Hbn.) female's oviposition behavior.