The significance of neuroplasticity in the respiratory control system has been appreciated only in recent years. One of the best studied models of respiratory plasticity to date is respiratory long-term facilitation (LTF), a progressive increase in respiratory motor output lasting several hours following acute intermittent hypoxia (AIH, 3 to 10 episodes), but not following sustained hypoxia of similar cumulative duration. Although the functional significance of LTF remains uncertain, suggested roles include stabilizing breathing during sleep or offsetting respiratory depression during hypoxia. However, regardless of its functional significance, our perspective is that the capacity to elicit LTF may be harnessed as a therapeutic approach for multiple clinical disorders of ventilatory control. Thus, a detailed understanding of cellular and synaptic mechanisms of LTF may provide the rationale for new pharmacological approaches in the treatment of severe ventilatory control disorders including obstructive sleep apnea or respiratory insufficiency during spinal cord injury or neurodegenerative motor neuron disease (e.g. ALS). Our understanding of the cellular/synaptic mechanisms that underlie LTF in phrenic motor output (pLTF) has increased dramatically in recent years. Our working model is that intermittent hypoxia triggers intermittent serotonin release near phrenic motoneurons, initiating pLTF by activating 5-HT2A receptors on their dendrites. Serotonin receptor activation initiates signaling cascades (e.g. PKC activation), stimulating new protein synthesis necessary for pLTF maintenance. New synthesis of brain-derived neurotrophic factor (BDNF) is necessary for pLTF since intrathecal administration of siRNAs targeting BDNF mRNA abolish increased BDNF concentrations in the ventral cervical spinal cord and pLTF. BDNF activates its high affinity receptor, TrkB, thereby phosphorylating and activating ERK MAP kinases and protein kinase B. We postulate that these kinases induce glutamate receptor trafficking, inserting more receptors in the post-synaptic membrane and strengthening synapses between (glutamatergic) brainstem respiratory neurons and phrenic motoneurons. Cellular pathways leading to pLTF are regulated by inhibitory constraints. For example, serine/threonine protein phosphatases appear to be an important regulator of pLTF expression. Intrathecal injections of okadaic acid (a potent inhibitor of multiple serine/threonine protein phosphatases) reveal pLTF following brief exposures to sustained hypoxia, a stimulus normally unable to engage this mechanism. On the other hand, intraspinal okadaic acid alone does not facilitate phrenic motor output, demonstrating that protein phosphatase inhibiton is necessary, but not sufficient, to induce pLTF. Since intrathecal okadaic acid has no effect on pLTF expression following AIH, we postulate that the protein phosphatase contraint of pLTF is diminished by AIH, and that differential regulation of protein phosphatase activity accounts in part for pLTF pattern-sensitivity. Differential inhibition of protein phosphatases by AIH versus sustained hypoxia may result form differential reactive oxygen species (ROS) formation in these conditions. Our hypothesis is that greater ROS formation during AIH (versus sustained hypoxia) enables pLTF through greater inhibition of relevant spinal phosphatases. Indeed, ROS are potent inhibitors of some serine/threonine protein phosphatases. In agreement, AIH-induced pLTF is impaired by administration of a superoxide dismutase mimetic or NADPH oxidase inhibitor, thereby demonstrating that ROS are necessary for pLTF expression. A link between the ROS and protein phosphatase inhibiton requirement of pLTF is provided by observations that intrathecal okadaic acid restores AIH-induced pLTF in rats treated with a superoxide dismutase mimetic. Thus, protein phosphatase inhibiton offsets the pLTF ROS requirement, an observation consistent with the idea that the primary actions of ROS in regulating pLTF are via their inhibitory actions on okadaic acid-sensitive phosphatases. ROS may also regulate other points in the cellular cascade of pLTF. Considerable progress is being made in understanding the fundamental mechanisms of LTF, including its inhibitory constraints. Such an understanding is critical if we are utilize this knowledge in the development of therapeutic approaches to the treatment of ventilatory control disorders. Strategies utilizing pharmacological agents, or new technologies such as RNA interference, may be most effective if we target molecules that, when inhibited, confer a gain of function.