1. This study examines the expression of beta 3-adrenoceptor messenger RNA (beta 3-AR mRNA) in rat tissues to allow comparison with atypical beta-adrenoceptors determined by functional and radioligand binding techniques. 2. A reverse transcription/polymerase chain reaction protocol has been developed for determining the relative amounts of beta 3-AR mRNA in rat tissues. 3. Measurement of adipsin and uncoupling protein (UCP) mRNA was used to examine all tissues for the presence of white and brown adipose tissue which may contribute beta 3-AR mRNA. 4. The beta 3-AR mRNA is expressed at high levels in brown and white adipose tissue, stomach fundus, the longitudinal/circular smooth muscle of both colon and ileum, and colon submucosa. There was substantial expression of adipsin in colon submucosa and moderate expression in fundus, suggesting that in these regions at least some of the beta 3-AR signal may be contributed by fat. Pylorus and colon mucosa showed moderate levels of beta 3-AR mRNA with lower levels of adipsin. Ileum mucosa and submucosa showed low but readily detectable levels of beta 3-AR. 5. Expression of adipsin in rat skeletal muscles coupled to very low levels of beta 3-AR mRNA indicates that the observed beta 3-AR may be due to the presence of intrinsic fat. beta 3-AR mRNA was virtually undetectable in heart, lung and liver. These results raise the possibility that the atypical beta-AR demonstrated by functional and/or binding studies in muscle and in heart is not the beta 3-AR. 6. By use of two different sets of primers for amplification of beta 3-AR cDNA, no evidence was found for differential splicing of the mRNA in any of the tissues examined. 7. The detection of beta 3-AR mRNA in the gut mucosa and submucosa suggests that in addition to its established roles in lipolysis, thermogenesis and regulation of gut motility beta 3-AR may subserve other functions in the gastrointestinal tract. The absence of beta 3-AR mRNA in rat heart or its presence with adipsin in skeletal muscle suggests that atypical beta-adrenoceptor responses in heart and skeletal muscle are unlikely to be mediated by beta 3-AR.