A truncated dopamine D3-receptor-like mRNA, named D3nf, predicts a protein that differs from the D3-receptor only in the carboxyl terminus. However, such a protein has lost the predicted membrane topology typically found for G protein-coupled receptors. Results presented here show that D3nf mRNA arises from the D3-encoded primary transcript via alternative splicing. This splicing, however, appears to involve cleavage of an unusual 3' splice site. Therefore, we tested the possibility that D3nf mRNA results from a splicing error. If this were the case, D3nf mRNA would be expected to be present in the cytoplasm only at very low amounts, and it would not be expected to be translated into protein. However, the relative abundance of cytoplasmic D3/D3nf mRNA in human cortical tissues was found to be similar. Furthermore, we raised polyclonal antisera against the predicted carboxyl-terminal peptide sequence of D3nf that reacts specifically with a protein expressed in stably D3nf mRNA-expressing COS 7 cells. The use of this antiserum also revealed the presence of a approximately 68 kDa D3nf-like immunoreactive protein in human brain, suggesting that the atypically processed D3nf mRNA is translated.