During studies on enteric co-innervation in the human esophagus, we found that not all acetylcholinesterase (AChE)-positive motor endplates stained for alpha-bungarotoxin (alpha-BT) and the vesicular acetylcholine transporter (VAChT), respectively. Therefore, we probed for differences in neuromuscular junctions in human esophagus by using triple staining for VAChT, alpha-BT, and AChE followed by qualitative and quantitative analysis. To exclude that the results were caused by processing artifacts, we additionally examined the influence of a number of factors including post-mortem changes and the type and duration of fixation on the staining results. Four types of neuromuscular junction could be distinguished in human esophagus: type I with VAChT-positive and type II with VAChT-negative nerve terminals on a alpha-BT-positive and AChE-positive endplate area, type III with VAChT-positive nerve terminals on a alpha-BT-negative but AChE-positive endplate area, and type IV with VAChT-negative nerve terminals on a alpha-BT-negative but AChE-positive endplate area. On average, 32% of evaluated AChE-positive motor endplates were type I, 6% type II, 24% type III, and 38% type IV. Based on these results, we suggest that, in human esophagus, (1) the most reliable method for staining motor endplates is presently AChE histochemistry, (2) alpha-BT-sensitive and alpha-BT-resistant nicotinic acetylcholine receptors exist in neuromuscular junctions, and (3) different types of VAChT or transport mechanisms for acetylcholine probably exist in neuromuscular junctions.