This thesis is concerned with the link between urban form and travel behaviour. We focus on two issues that have been overlooked in the existing empirical literature: the measurement of land use mix and the relation between urban structure and city-wide public transport patronage. In the last twenty years, there has been a substantial interest on the effects of land use mix on travel behaviour. The literature uses a variety of metrics to measure land use mix, but there is limited understanding of how the choice of metric affects estimates of the link between land use mix and travel. Researchers also measure urban form, and in particular land use mix, at various geographical scales. Past studies examining the effects of spatial scale on estimates of the urban form-travel demand relationship, offer limited evidence on land use mix measures. The first part of this thesis examines how the metric and spatial scale used in land use mix measurement affect our understanding of the relation between land use mix and travel, employing both simulation and empirical econometric analyses. The simulation analysis uses randomly generated data to construct and test alternative land use mix measures. The empirical analysis tests alternative metrics and alternative spatial scales in the context of public transport trip frequency in London. The last part of the thesis, presents some further econometric models of public transport demand, to investigate the effect of city structure on city-wide public transport patronage. The study extends previous work on the topic in a number of ways. These include considering more detailed measures of urban from; dealing with methodological issues related to endogeneity and experimenting with various functional forms.