We attempted to cultivate muscle cells from chick embryos in a serum-free, defined medium similar to that proposed by Bottenstein and Sato (1979) for the growth and differentiation of a murine neuronal cell-line. (1) We found that muscle cells from the legs of 11-day old chick embryos can be cultivated in a medium containing the different components indicated by Bottenstein and Sato, with 2 g/l bovine serum albumin, without serum or chick embryo extract. Myoblasts attached to the gelatin-coated dishes without any addition of attachment factors. They differentiated into myotubes in a similar manner as in classical serum supplemented media. (2) The level of cellular AchE activity was comparable in cultures grown in the presence of fetal calf serum (FCS), of horse serum (HS) and in the defined medium. The percentage of A(12) form was however higher in the defined medium (25-30%) than in FCS supplemented medium (about 5-6%). In HS supplemented medium the A(12) form was not detectable, partly because horse serum contains immunoglobulins which bind chicken AChE. The addition of defined medium components to FCS medium cultures did not lead to an increase of A(12). In contrast, the addition of a small amount (1%) of fetal calf serum to DM cultures reduced the level of A(12) in a drastic manner. FCS components therefore seem to repress the biosynthesis of A(12) AChE, or increase its degradation. (3) We estimated intracellular and extracellular compartments of AChE. The ratio of endocellular to ectocellular AChE decreased with the age of the cultures. The G(1) form was intracellular at all stages analyzed, but the other molecular forms were located in both cellular compartment, in different proportion: A(12) and G(4) seemed to be located preferentially in the external compartment, whereas G(2) was preferentially intracellular. (4) Muscle cultures grown in the defined medium and in the presence of serum secreted globular forms of AChE in a similar manner.