Purpose – The purpose of this study is to assess the evolution of restaurant locations in the city of Hamilton over a 12-year period (1996 to 2008) using GIS techniques. Retail theories such as central place, spatial interaction and principle of minimum differentiation are applied to the restaurant setting. Design/methodology/approach – A database of restaurants was compiled using the NZ yellow pages and contained 981 entries that consisted mainly of location addresses and types of cuisine. This paper focuses on locational patterns only. Findings – A process of geo-coding and clustering enabled the identification of two clustering periods over 12 years for city restaurants, indicating locational patterns of agglomeration within a short walking distance of the CBD and spill over effects to the north of the city. Research limitations/implications – The data do not allow statistical analysis of the variables causing the clustering but offer a visual description of the evolution. Explanations are offered on the possible planning regimes, retail provision and population changes that may explain this evolution. Practical implications – The findings allow identification of land use patterns in Hamilton city and potential areas where new restaurants could be developed. Also, the usefulness of geo-coded data in identifying clustering effects is highlighted. Originality/value – Existing location studies relate mostly to site selection criteria in the retailing industry while few have considered the evolution of restaurant locations in a specific geographic area. This paper offers a case study of Hamilton city and highlights the usefulness of GIS techniques in understanding locational patterns.