The focus of this paper is the analysis and prediction of local office rents and in particular the development of econometric models for two UK cities, Edinburgh and Glasgow. This paper reviews the current state of modelling and forecasting for office markets and notes the sparsity of urban office rent models. The urban models that exist suffer from data problems and either make the fatal flaw of ignoring supply constraints or consider supply in terms of net change in floorspace. The objective of this paper is to address some of the deficiencies in this empirical work on office market dynamics; it is the first attempt to use local take-up as a variable to model urban rents. Two approaches to modelling urban office rents are undertaken. The first model adopts a single reduced-form price equation using direct demand and supply measures, and suggests that variation in market dynamics exists between the two centres. However, these equations for the two cities have statistical weaknesses. The second model is a three-equation 'structural' model. The results of this analysis also suggest that Edinburgh responds more quickly to fundamental changes in supply-demand imbalances than Glasgow in the determination of office rents. The variation between Edinburgh and Glasgow, two cities within one administrative region of the UK, exemplifies the arguments in favour of urban analysis and the deficiencies in the regional approach to forecasting. The results of the empirical analysis also emphasise the importance of including local supply variables. Urban analysis still suffers from a paucity of published local economic data and recourse to regional data remains theoretically unsatisfactory. However, the use of a demand flow variable as demonstrated here is not only significant but arguably encompasses local economic drivers, and thereby negates to some extent the need for local economic indicators.