The N-allyl derivative (SK&F 85174) of 6-chloro-2,3,4,5-tetrahydro-1-(4-hydroxyphenyl)-1H-3-benzazepine-7,8-dio l (SK&F 82526) not only retains the exceptional D-1 agonist potency of its parent but also displays reasonably potent D-2 agonist activity, as measured by a dopamine-sensitive adenylate cyclase test and a rabbit ear artery assay, respectively. Several additional N-substituted compounds were prepared to explore the D-2/D-1 agonist relationship. The N-methyl analogue retained good D-2 agonist potency, but this substitution converted D-1 agonist activity into antagonist activity. Most other N-substituents sharply decreased D-2 agonist potency including the N-n-propyl group. This observation was surprising since the introduction of mono- or di-N-n-propyl substituent(s) is commonly linked with retention or enhancement of D-2 agonist potency in other series of dopamine agonists. The N-(2-hydroxyethyl) analogue retains about one-fourth the D-2 potency of SK&F 85174. Several synthetic methods were used to prepare these compounds. N-Allylation of a trimethoxybenzazepine followed by cleavage of the methyl ethers with boron tribromide was the preferred method. Other methods used were direct alkylation of the trihydroxy secondary amine, i.e., SK&F 82526, and an acylation-amide reduction-cleavage method.