1. Functional and molecular approaches were used to characterize the beta-AR subtypes mediating relaxation of rat ileal smooth muscle. 2. In functional studies, (-)-isoprenaline relaxation was unchanged by CGP20712A (beta1-AR antagonist) or ICI118551 (beta2-AR antagonist) but shifted by propranolol (pKB=6.69). (+/-)-Cyanopindolol, CGP12177 and ICID7114 did not cause relaxation but antagonized (-)-isoprenaline relaxation. 3. BRL37344 (beta3-AR agonist) caused biphasic relaxation. The high affinity component was shifted with low affinity by propranolol, (+/-)-cyanopindolol, tertatolol and alprenolol. CL316243 (beta3-AR agonist) relaxation was unaffected by CGP20712A or ICI118551 but blocked by SR58894A (beta3-AR antagonist; pA2 = 7.80). Enhanced relaxation after exposure to forskolin and pertussis toxin showed that beta3-AR relaxation can be altered by manipulation of components of the adenylate cyclase signalling pathway. 4. The beta-AR agonist RO363 relaxed the ileum (pEC50=6.18) and was blocked by CGP20712A. Relaxation by the beta2-AR agonist zinterol (pEC50=5.71) was blocked by SR58894A but not by ICI118551. 5. In rat ileum, beta1-, beta2- and beta3-AR mRNA was detected. Comparison of tissues showed that beta3-AR mRNA expression was greatest in WAT>colon=ileum >cerebral cortex>soleus; beta1-AR mRNA was most abundant in cerebral cortex > WAT > ileum = colon > soleus; beta2-AR mRNA was expressed in soleus > WAT > ileum = colon > cerebral cortex. 6. These results show that beta3-ARs are the predominant beta-AR subtype mediating rat ileal relaxation while beta1-ARs may produce a small relaxation. The beta2-AR agonist zinterol produces relaxation through beta3-ARs and there was no evidence for the involvement of beta2-ARs in relaxation despite the detection of beta2-AR mRNA.