Fresh water scarcity is an increasing problem worldwide. Strategies to alleviate water scarcity include the use of low-quality water for irrigation. The risk of groundwater contamination by pollutants in this water is affected by soil heterogeneity and preferential flow. These risk factors can be assessed by measuring the spatio-temporal redistribution of uniformly applied water and solutes. We placed a soil monolith (height 29 cm) from an Australian vineyard on a 100-cell multi-compartment sampler (MCS). At this vineyard, treated wastewater is used in response to the severe shortage of water in the summer. We studied the leaching risk associated with heterogeneous or preferential flow by irrigating the soil column with 24 applications to simulate one year. We applied simulated rainfall as well as wastewater (which contained chloride) during summer while relying on rainfall only in winter. We compared the chloride leaching with the leaching of bromide, which was applied during one of the applications as a pulse. During the entire simulated year, leaching of solutes from the monolith was measured. The results indicate that the assumption of uniform flow would underestimate the risk for the fresh groundwater reserves: 25% of the solutes are transported though 6% of the soil's cross-section. The spatial distribution of drainage and solute leaching varied little during the experiment. Consequently, the mass flux density pattern of the bromide pulse was comparable to that of the repeatedly applied chloride. However, the MCS data suggested lateral 'escape' from chloride to non-mobile areas, which means in the long run, considerable quantities of these solutes can build up in areas that do not receive irrigation water.