Vascular smooth-muscle cells (VSMCs) from rat aortae contained very little non-muscle myosin heavy chain (MHC) immediately after dispersal, and the protein did not accumulate if the cells were held in G0/G1 phase by withholding serum or were held in first S phase by the addition of bromodeoxyuridine (BrdU). However, non-muscle MHC accumulated by greater than 20-fold per cell during first M phase, when over 80% of the cells divided between 48 h and 72 h after addition of serum. Delaying the addition of serum caused a delay in the accumulation of the non-muscle MHC until the cells subsequently entered M phase. If the cells were held in M phase at the metaphase/anaphase boundary by nocadazole, the accumulation of non-muscle myosin still occurred, although division was blocked. When the cells were pulse-labelled with [35S]methionine, it was found that non-muscle MHC was one of the major proteins being made and that its synthesis occurred at similar rates throughout the cell cycle. This implied that the rate of degradation of the protein before first M phase was much faster than in M phase, when the protein accumulated rapidly. This was confirmed by direct measurements of the rate at which [35S]methionine-labelled non-muscle MHC disappeared from the cells, which gave a half-life for the protein of about 8 h before M phase but about 5 days in post-mitotic cells, i.e. an increase of approx. 15-fold. These data are consistent with the hypothesis that there is a mechanism in VSMCs which shortens the half-life of the protein before first M phase and that the accumulation of non-muscle MHC which results from the increase in half-life at first M phase may be necessary for division of these cells.