Friend murine erythroleukemia (F-MEL) cells were transfected with a plasmid bearing tandemly arranged mouse c-myc antisense and dihydrofolate reductase transcription units. Sixteen clones were isolated, each containing unrearranged c-myc sequences and expressing high levels of antisense transcripts. All antisense clones examined contained reduced amounts of cytoplasmic endogenous c-myc transcripts. The kinetics of reaccumulation of endogenous c-myc mRNA during a 24-h exposure to dimethyl sulfoxide (DMSO) were also retarded and the ultimate transcript levels attained were less than in control cells. Antisense clones grew as well as control F-MEL cells in medium containing 10% fetal calf serum but at only a half and a quarter of the control rates in media containing 5 and 2% serum, respectively. Antisense clones differentiated faster and to a greater degree than control cells following DMSO exposure. myc antisense transcript expression was increased by growing cells in methotrexate, which resulted in an enhanced response to DMSO. Fluorescence-activated cell sorter (FACS) analysis of cellular DNA content indicated that a greater fraction of antisense nuclei contained a G₀/G₁ 2n DNA content following a 24-h exposure to DMSO. When density-arrested antisense clones were diluted into fresh medium to allow reentry into the cell cycle, they incorporated less [3H]thymidine than control cells. FACS analysis showed that this was because only a portion of the cell population was entering S phase. Whereas control cells did not increase in size following release from density arrested antisense cells contained a subpopulation which were initially smaller and which eventually attained the same size as control cells. Quiescent antisense cells thus comprise two populations, each arrested at a different point in G₁. Dilutional replating allowed both populations to reenter the cell cycle. We propose a model which postulates that certain minimal myc levels are necessary for cells to traverse G₁. Those with insufficient levels, due, for example, to antisense inhibition, are unable to completely traverse G₁ during density arrest and synchronize at an earlier point than do control cells. This earlier point may be along the differentiation pathway and may account for the greater responsiveness of antisense cells to DMSO induction. This model postulates that F-MEL cells overexpressing myc fail to differentiate because myc levels are never sufficiently low enough to allow cells to enter the differentiation pathway.