The genes of an individual are said to be preserved in his relatives i
f they possess, collectively, all copies of his genes, We present a me
thod for computing the probability that an individual's genes are pres
erved in his relatives, Using this method, we compute gene preservatio
n probabilities (up to three linked loci) for a variety of relationshi
ps for humans and for haplodiploid species. The results suggest that s
ome widely held notions in ethology and sociobiology seem to be questi
onable, From the gene preservation viewpoint, two brothers are far fro
m enough to justify the sacrifice of one's own life, because the proba
bility that an individual's genes are preserved in his two siblings is
dismally small, The precise probability that an individual's genes ar
e all preserved in a group of his relatives depends on the precise spe
cification of the pedigree structure. We also demonstrate that, for a
hymenopteran female, there is no practical difference, in terms of gen
e preservation probability, between helping her sisters to breed and b
reeding her own offspring. In fact, since the genes of her sister will
be either lost or preserved in her nieces/nephews, it is more appropr
iate to compare the probability of preserving her genes through her ow
n offspring with that through her nieces/nephews. We show that her cha
nce of preserving all her genes is much higher if she chooses to breed
her own offspring instead of helping her sister.