The invasion of sewer pipes by tree roots is a major cost to both the corporations managing urban infrastructure and to private landowners. There are a number of factors that may result in the growth of roots into and around the pipes. These include pipe material factors such as degradation of aging pipes and damage caused to pipes and environmental factors such as site conditions, tree proximity and tree species. Research into these issues will help in developing more efficient methods of controlling root invasion, benefiting both the management of sewer flow and the urban landscape. A study of root invasions in suburban Melbourne was carried out to examine relationships between site factors and root blockages. Common factors found associated with tree root blockages were Eucalyptus and Melaleuca species over 4m high within 6 m of sewers, pipe depth 2 m or less, rubber ring joined 150 mm diameter vitreous clay, 30 to 60 years old, soil types with sandy topsoils, and blockages occurring most frequently when temperatures and evaporation were at their lowest i.e. August to October. An experiment was conducted to evaluate chemical and physical treatments for preventing root growth in disturbed soil with high nutrient content and readily available moisture (modeling leaking sewers). Over a period of 177 days significant inhibition was achieved with the chemical treatments dichlobenil (278 g/m and 1392 g/m), copper sulphate (1.5 g Cu/kg soil and 7.5 gCu/kg soil), oryzalin (1031 g/m) and trifluralin (260 g/m) and the physical treatments slaked lime (ratio of 10 soil:1 lime and 10 soil:5 lime) and cement slurry. The potential for the use of these treatments in areas where tree roots have damaged sewers is discussed. INTRODUCTION The degree of harmony between trees and urban infrastructure depends on the levels of understanding of how, why and when tree root damage arises and what prevention strategies are available. Tree impacts on infrastructure such as sewers can be a major cost to the managers of utilities and other urban infrastructure managers (Schrock 1985; Baker, 1980; McPherson and Peper 1996). Soil disturbance during installation creates pathways for roots to grow from the surface towards the pipe. Essentially, these root growth corridors contain favourable sized pores that can hold water and air, similar those found in topsoil. If pipes have been damaged and leak, then root growth will eventually exploit such regions. Tree roots, incorrect installation procedures, other underground service operators or material degradation, may all cause damage. This paper reports on a study, which investigated the factors associated with sewer blockage in metropolitan Melbourne and which investigated techniques that could be used to minimise tree root growth in disturbed soils such as found near sewer pipes. MATERIALS AND METHODS The survey was conducted in two parts. The first was an investigation of site factors associated with tree root blockages in the year 1999 by South East Water (SEW). Proc. IC on Urban Horticulture Eds: R. Junge-Berberovic et al. Acta Hort 643, ISHS 2004 114 Sites where blockages had occurred were visited and specific information about the site recorded. The second part was an analysis of factors associated with tree root blockages that occurred in the four years 1997 to 2000. This component of the study did not involve site visits but it used the available information on blockages recorded by SEW. The study of 1999 blockages recorded data on trees, geology and soil. The study of 1997-2000 blockages examined a range of factors that may have influenced blockages. In December 1999, a site with 85 trees planted in four rows was selected at Burnley College. The trees were Eucalyptus globulus, Eucalyptus saligna, Eucalyptus camaldulensis, Eucalyptus melliodora and Populus canescens. Holes were dug the trees and then soil sections were then were backfilled with treated sandy loam soil. The treated soil section was watered and fertilised periodically to simulate a soil that is situated close to a leaking pipe. The concept of this experiment was based on studies by Nazer and Clark (1982) and Wagar and Barker (1993). There were twelve treatments (nine chemical, two physical, and one control (no treatment)). The treated cores were excavated after 177 days and the root mass in each was measured. The treatments used are displayed in Table 1. The cement slurry and compaction are categorised as ‘physical treatments’ because they act as a physical barrier to root growth. There were five replicates of each treatment.