Brain regions typically contain intermixed subpopulations of neurons with different connectivity and neurotransmitters. This complicates identification of neuronal phenotypes in electrophysiological experiments without using direct detection of unique molecular markers. A prime example of this difficulty is the identification of dopamine (DA) neurons in the midbrain ventral tegmental area (VTA). Although immunocytochemistry (ICC) against tyrosine hydroxylase (TH) is widely used to identify DA neurons, a high false negative rate for TH ICC following ex vivo electrophysiology experiments was recently reported, calling into question the validity of comparing DA and non-DA VTA neurons based on post-hoc ICC. However, in whole cell recordings from randomly selected rat VTA neurons we have found that TH labeling is consistently detected in ∼55% of neurons even after long recording durations (range: 2.5-150 min). This is consistent with our prior anatomical finding that 55% of VTA neurons are TH(+). To directly estimate a false negative rate for our ICC method we recorded VTA neurons from mice in which EGFP production is driven by the TH promoter. All 12 EGFP(+) neurons recorded with a K-gluconate internal solution (as used in our rat recordings) were strongly labeled by TH ICC (recording duration 16.6±1.8 min). However, using recording electrodes with an internal solution with high Cl(-) concentration reduced the intensity of TH co-labeling, in some cases to background (recording duration 16.7±0.9 min; n = 10). Thus TH is a highly reliable molecular marker for DA neurons in VTA patch clamp recordings provided compatible microelectrode solutions are used.