The importance of heterosis, and in particular allozyme-associated heterosis, in natural populations remains unclear. Much of the scepticism that surrounds allozyme-associated heterosis comes from inconsistent and unreliable detection of the relationship. Thus, evaluating the genetic mechanisms that may cause allozyme-associated heterosis in natural populations has proven difficult. The most prevalent hypotheses that have been put forward to explain the genetic basis of heterosis are the general dominance and the local overdominance hypotheses. A factorial crossing design was used to survey eight polymorphic allozyme loci in the parent and offspring generations of two species of earthworms in order to evaluate possible mechanisms of allozyme-associated heterosis for growth rate. Significant heritable variation for growth rate was detected only within a single cross. Allozyme-associated heterosis for growth rate was detected only within this cross. This relationship did not persist after the effects of interfamily variation were removed. These results indicate that simple heritability of a fitness-related trait may be necessary for predictive power and repeatability of allozyme-associated heterosis, and that the allozyme-associated heterosis detected in this study was the result of general dominant genetic effects.