To locate the sites of dopamine (D) production in rat renal cortex, we separated glomeruli and proximal tubules by sieving or centrifugation in Percoll after collagenase digestion. After centrifugation layer I contained 60–80% glomeruli and 20–40% tubule fragments, half of which did not stain with alkaline phosphatase, layer II contained 0–5% glomeruli, 10–25% tubule fragments other than proximal tubules, and 70–85% proximal tubule fragments. Layer IV contained 85–95% proximal tubules. Gluconeogenic rates were (micromoles per hour per gram wet weight) as follows: I, 4 ± 1; II, 7 ± 2; and IV, 16 ± 1. Norepinephrine (NE) content was (picomoles per gram wet weight) I, 310 ± 30; II, 540 ± 40; IV, 195 ± 60. D content was (picomoles per gram wet weight) I, 26 ± 6; II, 46 ± 13; IV, 33 ± 7. Surgical denervation 4–6 days previously reduced the norepinephrine content of layers I and II to 35 ± 10 (p < 0.001) and of IV to 60 ± 20 (p < 0.05) and the D content of layers I and II to 13 ± 6 and 6 ± 6 pmol/g, respectively (p < 0.01); D content of layer IV was unchanged. D production from 10−7 M 3,4-dihydroxyphenylalanine (DOPA) was (nanomoles per gram per minute) I, 0.2 ± 0.03; II, 0.7 ± 0.1; IV, 1.0 ± 0.04. DOPA consumption was (nanomoles per gram per minute) I, 0.6 ± 0.1; II, 1.4 ± 0.3; and IV, 1.8 ± 0.2. Denervation did not change D production or DOPA consumption. Glomeruli without arterioles or tubules, obtained by sieving, contained no D or NE and did not produce D from DOPA. We conclude that proximal tubules produced at least five times more D from DOPA than glomeruli and distal tubules. Renal nerves did not contribute significantly to D production. We found no evidence of specific D-containing nerves associated with rat glomeruli.