The complete RNA genome of turnip mosaic potyvirus (TuMV) was amplified by seven consecutive reverse transcriptase-polymerase chain reactions and cloned into pUC9. The viral RNA is 9830 nucleotides long and contains a single open reading frame (ORF) of 9489 bases encoding a large polyprotein of 3863 amino acids with a calculated M r of 358000. The non-coding region (NCR) preceding the ORF is 129 nucleotides long and has a high AU content (70%). Its predicted secondary structure is characterized by a hairpin loop with a free energy loss of -69.9 kJ/mol. The termination codon is followed by an AU-rich NCR of 209 bases, excluding the poly(A) tail. Seven potential nuclear inclusion a proteinase (NIa-Pro) recognition heptapeptides are found in the polyprotein. Their sequences agree with consensus potyviral NIa-Pro cleavage sequences except for that at the 6K-VPg site, which is characterized by a glutamic acid residue preceding the hydrolysed peptide bond. The TuMV proteins are similar to their corresponding potyviral proteins.