Population size, structure and demographic history are central, complicating factors in prediction of inbreeding depression, the decline in fitness commonly seen with increasing homozygosity. The effects of inbreeding have been extensively studied, largely theoretically, and supporting data within and across species has not always conformed to expectation. We made highly replicated measurements of inbreeding depression for multiple isolates of globally distributed Caenorhabditis species. Selfing has evolved repeatedly in this genus, yet inbreeding depression is extremely strong in outcrossing species and it is unclear how the short-term benefits of reproductive assurance have been accessed. From genomes assembled and annotated de novo, we estimated effective population size, and fixed and segregating loads from variant impact prediction. Some of these estimates are highly predictive of inbreeding depression across morphologically similar, yet deeply diverged, species.