PurposeThis paper aims to build a set of long-term, geographically controlled land value indices for Santiago de Chile, with which to test land rent theory predictions regarding macroeconomic impacts.Design/methodology/approachThis paper uses a geographic cluster approach to the Laspeyres estimator, weighted by the stock of available land plots and their market offers per zone, to create two quarterly land value indices for Gran Santiago during the period 1983Q4-2016Q2. Subsequently, this paper implements dynamic time series methods (Vector Error Correction) as a baseline to determine the effect of economic performance and interest rate on urban land values.FindingsThe two land value indices are correctly predicted by economic and interest rate shocks, as theoretically expected. In addition, this paper found that land values grew faster-than-predicted during the period of the so-called "Chilean Miracle" (1992-1998), a situation associated in the literature with worsened housing affordability and socio-spatial inequality.Originality/valueThe land value indices offer unprecedented time-series precision in tracking the long-term performance of real estate markets in a Latin American city, allowing us to produce short- and long-term (accumulative) time-series causality analyses.