Primary visual (V1), auditory (A1) and somatosensory (S1) cortices contain well-ordered maps of the sensory periphery. The retinotopic (V1), tonotopic (A1) and somatotopic (S1) maps are largely determined through the arrangement of incoming thalamocortical projections. Whereas it is clear that activity-independent (Molnar et al., 2002) genetically-programmed mechanisms involving molecular gradients (Fukuchi-Shimogori & Grove, 2001) play a major role in the establishment of these thalamocortical maps during development, there is also strong evidence indicating that activity contributes substantially to cortical map refinement. A convenient model system for investigating cortical map formation and plasticity is the mouse primary somatosensory barrel cortex (recently reviewed by Petersen, 2007). In this cortical region, each mystacial vibrissa on the snout of the rodent is represented by an anatomically defined structure in layer 4, termed a 'barrel', and the resulting somatotopic barrel map can be visualised through a large number of simple stains.In this issue, She et al. (2009) find that the barrel map is partially disrupted in metabotropic glutamate receptor type 5 (mGluR5) knockout mice, in good agreement with a previous study (Hannan et al., 2001). Of particular interest are alterations in the representation of the large posterior vibrissae that are actively swept backwards and forwards during whisking behaviour in exploring mice. For these whiskers, the thalamocortical afferents in mGluR5 knockout mice segregate normally into clusters in layer 4, a process which is thought to be a key event driving barrel formation. However, the organisation of postsynaptic neurons in layer 4 is disturbed in knockout mice. In wild type and heterozygous mice, the somata of layer 4 neurons form cell-dense walls surrounding each of the thalamocortical axon clusters, but this pattern is completely absent in mGluR5 knockout mice. This clearly indicates that there are at least two separable steps for barrel formation, one being the segregation of the presynaptic thalamocortical axons (which appears to be normal in mGluR5 knockout mice) and a second step to position the layer 4 cell bodies into a barrel map (which is disrupted in mGluR5 knockout mice).In wild type and heterozygous mice, the neurons located in barrel walls exhibit highly asymmetric dendrites pointing towards the barrel center, which presumably contribute to the sharpening of receptive fields. Such oriented dendrites were found much less frequently in mGluR5 knockouts. Perhaps resulting from these unpolarised neurons sending their dendrites into more than one thalamocortical cluster, She et al. (2009) find unusual short latency responses to deflection of more than one whisker in the knockout mice.She et al. (2009) also studied thalamocortical synaptic transmission and plasticity in vitro revealing several differences in mGluR5 knockout mice. Although presynaptic thalamocortical function and AMPA-receptor-dependent signalling appeared normal, postsynaptic differences ...