Mar. Ecol. Prog. Ser., vol. 156, 225-237 In order to understand the phenomenon of heterozygote deficiency (F is) in marine molluscs, all the relevant literature concerning this phenomenon in the Mytilus edulis species complex was reviewed.Due to large heterogenity in methods of data analysis, in particular for the choice of the estimator measuring heterozygote deficiency and for the choice of the testing procedure, no clear overview was possible. To overcome this problem, similar estimators and tests were used to analyze or re-analyze some additional large data sets from the Baltic (M. trossulus), France (M. galloprovincialis), England (M. edulis), Atlantic USA (M. edulis) and Pacific USA (M. trossulus). Large end significant heterozygote defincies exist in these data sets. Estimated F is (F is) values are generally higher for Lap than to other loci, consistent with an effect of selection. No other significant variation of F is values across loci within each data set or across data sets for each locus was detected, however no specific test has been designed for this null hypothesis. The possible contribution af a Wahlund effect to explain the heterozygote deficiency is discussed (such as a Wahlund effect or selection) in M. edulis or in other organisms, and the species-specific, locus-specific or population-specific explanations are to be sought. Affiliation: University of Montpellier F-34095 Montpellier cedex 05 France