Previously we have shown that leukaemia inhibitory factor (LIF) potentiates the development of murine spinal cord neurons in vitro, suggesting that it, or related factors, may play an important regulatory role in neuronal development. We have further investigated this role and show here that the generation of neurons in cultures of embryonic day 10 spinal cord cells is inhibited by antibodies to the beta subunit of the LIF receptor. Since there are more undifferentiated precursors in antibody-treated cultures than in control and LIF-treated cultures, it is concluded that the primary action of LIF, or related molecules, is to promote neuronal differentiation, not precursor survival. In addition, the failure of LIF to support neuronal survival in the period immediately following differentiation suggests that the increased numbers of neurons generated with LIF are not attributable to its neurotrophic action. By selecting neuronal precursors on the basis of their inability to express class 1 major histocompatibility complex molecules, it was shown that LIF acted directly upon these cells and not via an intermediary cell. LIF also appears to be involved in regulating the differentiation of astrocytes, since it increases the number of glial fibrillary protein (GFAP)-positive cells present in the cultures and since the spontaneous production of GFAP-positive cells is blocked by antibodies to the LIF beta receptor. These findings suggest that LIF or related factors promote the differentiation of neural precursors in the spinal cord, but that they are not involved in preferentially promoting precursors down a specific differentiation pathway.