iPSC-derived motor neurons (iPSC-MNs) from MND patients display disease phenotype in-vitro1 , but represent only one part of the neuromuscular partnership in the disease process. In-vitro neuromuscular junction (NMJ) interaction and pathology using human iPSC-MNs and muscle is a more clinically relevant platform to research the disease2 but NMJs are difficult to reproduce in standard co-culture systems. We hypothesise this reflects different cell-type media requirements. We are therefore developing microfluidic culture systems which allow precise compartmentalized environmental control, modelling the in-vivo segregation of neurons and muscle cell microenvironments3, and facilitating high resolution studies of disease processes and screening of potential therapeutics. Microfluidic chips (40±5 x 5 x 20±5 mm), were fabricated from moulded PDMS applied to glass coverslips (22 x 60 mm)4,5 , providing two parallel chambers (300 x 400 x 10 µm), with fine connecting cross-channels (400 x 4 x 4 µm). Human iPSC-MNs from healthy and MND patient donors (C9ORF72 repeat expansion), plus isogenic gene-corrected controls were cultured in neuronal chambers, and axons grew through cross-channels into a parallel muscle chamber. iPSC-MN marker expression was probed by immunofluorescent staining for HB9 and ChAT, quantified using mean fluorescence intensity. All data is mean±SEM. Neurons were visualised by immunofluorescence for Neurofilament-heavy (axons), MAP2 (dendrites) and Hoechst 33341 (nuclei), and C2C12 myoblasts with MitoTracker Red (mitochondria), α-bungarotoxin (acetylcholine receptors, AChRs), myosin (cytoskeletal). All iPSC-MN lines were cultured for 120 days and expressed mature MN markers. Axon length per cell increased over time in all chips (e.g. in healthy iPSC-MNs from 481.9±83.6 µm day 15 to 916.5±368.1 µm on day 120), with no significant difference between healthy, ALS or controls (n=3, p=0.683, two way ANOVA). In other experiments, C2C12 myoblasts were successfully differentiated to myotubes in-chip, with mean fluorescence intensity and myotube length calculated to assess differentiation (e.g. length of C2C12s at day 11, 119.0±60.6 µm; n=3). Finally, iPSCs-MNs and C2C12 myocytes were cultured in separate parallel compartments linked by micro-channels. iPSC-MN axon terminals grew towards C2C12 myotubes. While the percentage of AChR clusters associated with axon terminals in ALS was less (28.3±9.7%) than gene corrected controls (51.2±4.4%), this failed to reach significance (n=7 chips, p=0.0539, unpaired t-test). We have demonstrated successful culture and differentiation of human iPSC-MNs and an accessible muscle cell line in a novel design of microfluidic chips, as well as viable co-cultures that expressed markers of NMJ formation. This provides the basis of further motor neuron-muscle co-cultures to undertake high resolution studies of the physiology and pathology of MND in the different cellular components: MNs, muscle and the NMJ.