The molecular basis for leptospirosis remains poorly understood, with no efficient methods available for screening libraries of mutants for attenuation. We have analysed the attenuation of leptospiral transposon mutants in vivo using a high-throughput method by infecting animals with pooled sets of transposon mutants. A total of 95 mutants was analysed by this method in the hamster model of acute infection, and one mutant was identified as attenuated (M1233, lb058 mutant). All virulence factors identified in Leptospira to date have been characterised in the acute model of infection, neglecting the carrier host. To address this, a BALB/c mouse colonisation model was established. The lb058 mutant and two mutants defective in LPS synthesis were colonisation-deficient in the mouse model. By applying the high-throughput screening method a further five colonisation-deficient mutants were identified for the mouse model; these included two mutants in genes encoding proteins with a predicted role in iron uptake (LB191/HbpA and LB194). Two attenuated mutants had transposon insertions in either la0589 or la2786 (encoding proteins of unknown function). The final attenuated mutant had an unexpected deletion of five genes (la0969-la0975) at the point of transposon insertion. This is the first description of defined, colonisation-deficient mutants in a carrier host for Leptospira. These mutants were either not attenuated, or only weakly attenuated, in the hamster model of acute leptospirosis, thus illustrating different factors that may be required in the carrier and acute models of leptospiral infection. High-throughput screening can reduce the number of animals used in virulence studies and increase the capacity to screen mutants for attenuation, thereby enhancing the likelihood of detecting unique virulence factors. A comparison of virulence factors required in the carrier and acute models of infection will help to unravel colonisation and dissemination mechanisms of leptospirosis.