Films composed of nanotube networks have their conductivities regulated by the junction resistances formed between tubes. Conductivity values are enhanced by lower junction resistances but should reach a maximum that is limited by the network morphology. By considering ideal ballistic-like contacts between nanotubes, we use the Kubo formalism to calculate the upper bound for the conductivity of such films and show how it depends on the nanotube concentration as well as on their aspect ratio. Highest measured conductivities reported so far are approaching this limiting value, suggesting that further progress lies with metallic nanowires rather than carbon nanotubes. © 2009 American Institute of Physics. ͓doi:10.1063/1.3236534͔The search for thin films that are flexible, transparent, and conductive is driven by their potential as transparent electrodes. [1][2][3] One common route is to use films made by a disordered network of carbon nanotubes ͑NTs͒. [4][5][6][7][8][9] In this case, electrons move across the entire film by moving between NT in close proximity. The conductivity is limited by the tunneling between tubes, which introduces a significant inter-NT junction resistance. To make the films more conductive, one needs to improve the coupling between NT, which recently has been achieved with acid treatments. 10,11 Further attempts are being made to lower the junction resistance and surpass the best conductivity reported so far, which currently stands at Ϸ 6 ϫ 10 5 S / m. 5,10,11 In addition to the junction resistance, the network morphology also plays a role in limiting the film conductivity. In fact, we have recently demonstrated how sensitive to the network connectivity the conductivity can be. 12 This means that no matter how much progress is made in lowering the junction resistance, there should be a maximum value for the film conductivity, which is regulated by the network itself. This is the goal of the present manuscript, i.e., to obtain an upper bound for the conductivity of disordered NT networks. The knowledge of this upper bound should avoid overoptimistic expectations for the transport properties of the films. Furthermore, understanding the interplay between network morphology and the intrinsic conductances of NT may be explored to deal with films made of other nanowires ͑NWs͒. 13,14 Since we are interested in the best-case scenario in which the electronic conductivity is at its maximum, we must eliminate potential sources of scattering and decoherence such as structural imperfections, impurities, and interaction with other quasiparticles. In this situation, it is appropriate to consider a purely ballistic regime of transport within the NW, which calls for a quantum description of the conductivity. NTs are known to behave as ballistic conductors with two quanta of conductance 15 and are often referred to as possessing two conducting channels. How much interference there is between these channels is what determines how the network affects the film conductivity. Furthermore, because other quantum wir...