The kinship theory of genomic imprinting proposes that parent-specific gene expression evolves at a locus because a gene's level of expression in one individual has fitness effects on other individuals who have different probabilities of carrying the maternal and paternal alleles of the individual in which the gene is expressed. Therefore, natural selection favors different levels of expression depending on an allele's sex-of-origin in the previous generation. This review considers the strength of evidence in support of this hypothesis for imprinted genes in four "clusters," associated with the imprinted loci Igf2, Igf2r, callipyge, and Gnas. The clusters associated with Igf2 and Igf2r both contain paternally expressed transcripts that act as enhancers of prenatal growth and maternally expressed transcripts that act as inhibitors of prenatal growth. This is consistent with predictions of the kinship theory. However, the clusters also contain imprinted genes whose phenotypes as yet remain unexplained by the theory. The principal effects of imprinted genes in the callipyge and Gnas clusters appear to involve lipid and energy metabolism. The kinship theory predicts that maternally expressed transcripts will favor higher levels of nonshivering thermogenesis (NST) in brown adipose tissue (BAT) of animals that huddle for warmth as offspring. The phenotypes of reciprocal heterozygotes for Gnas knockouts provide provisional support for this hypothesis, as does some evidence from other imprinted genes (albeit more tentatively). The diverse effects of imprinted genes on the development of white adipose tissue (WAT) have so far defied a unifying hypothesis in terms of the kinship theory.