Several recent theoretical considerations of mating-sqstem evolution p
redict within-population covariation between levels of inbreeding depr
ession and genetically controlled mating-system characters. If inbreed
ing depression is caused by deleterious recessive alleles, families wi
th characters that promote self-pollination should show lower levels o
f inbreeding depression than families with characters that promote out
crossing. The converse is expected if inbreeding depression is due to
overdominant allelic interactions. Whether these associations between
mating-system and viability loci evolve will have important consequenc
es for mating-system evolution. The evolution of selfing within the ge
nus Mimulus is associated with a reduction in stigma-anther separation
(i.e., a loss of herkogamy) and high autogamous seed set. In this stu
dy we compared families from two M. guttatus populations that differed
genetically in their degree of stigma-anther separation. In one of th
ese populations we also compared families that differed genetically in
the degree to which they autogamously set seed in a pollinator-free g
reenhouse. Dams often differed significantly in levels of inbreeding d
epression for aboveground biomass and flower production, but variation
in inbreeding depression was never explained by herkogamy class or au
togamy class. Several factors might account for why families with trai
ts associated with selfing did not show lower inbreeding depression, a
nd these are discussed. Our study also demonstrated significant variat
ion among self progeny from a given female likely due to differences i
n pollination date and position of fruit maturation. The detection of
significant dam X sire interactions suggests biparental inbreeding or
differences in combining ability for specific pairs of parents.