Maintenance of existing road network represents a challenge for public road authorities who seek a balance between available budgets and the need for maintaining level of service at a satisfactory level on existing road sections. For this reason, prediction of cost for road rehabilitation and reconstruction works represents one of key inputs for the objective analysis of projects and available budgets and optimization of road maintenance alternatives.
However, the average unit costs of road rehabilitation and reconstruction vary substantially between countries, and even between projects in the same country, due to a number of factors. In this paper an effort is made to develop a prediction model that could be applied for a wide range of conditions in different countries. A specialized dataset is used, which was generated under a World Bank study for a sample of road works contracts from 14 countries in Europe and Central Asia, signed between the years 2000 and 2010. The data sample for the analysis covers 94 projects of rehabilitation and reconstruction of flexible pavements. A multivariate regression analysis is used to evaluate the determinants of the cost per kilometer of the road rehabilitation or reconstruction.
The explanatory variables that are tested in the model are divided in three groups:
a Variables related to oil prices
b Variables that are country specific
c Variables that are project specific
The variables included in the analyses were chosen in view of their potential explanatory power. The resulting regression model is expected to be useful in the strategic analysis of road networks, including the optimization of road maintenance alternatives.

Road rehabilitation and reconstruction works, cost prediction model, multivariate linear regression analysis