About half of all angiosperms have some form of molecular self-incompatibility to promote outcrossing. If self-incompatibility breaks down, inbreeding depression (δ) is the main barrier to the evolution of self-fertilisation (selfing). If inbreeding depression is lower than 50% (δ < 0.5), the inherent transmission advantage of selfers should theoretically drive the evolution of selfing. However, this does not always happen in practice. For example, despite frequent breakdowns of self-incompatibility in North American Arabidopsis lyrata, selfing has only evolved in few populations. This is surprising given that previous inbreeding-depression estimates were well below the 0.5 threshold. Here, we test whether this could be due to underestimation of true inbreeding depression in competition-free environments. Specifically, we tested whether direct competition between crossed and selfed siblings magnified inbreeding-depression estimates in A. lyrata. We found that this was neither the case for belowground nor for aboveground biomass. For reproductive traits, there was hardly any significant inbreeding depression regardless of competition. Combined with previous findings that drought stress and inducing defence also did not magnify inbreeding depression, our results suggest that the relatively low estimates of inbreeding depression for biomass are indeed realistic estimates of the true inbreeding depression in North American A. lyrata.