The contents of heavy metals in soil and vegetable samples collected from an urban garden in Kayseri, Turkey, were investigated. Both wet- and dry-ashing methods were used for dissolving vegetable samples. A sequential extraction procedure proposed by the Commission of the European Communities, Community Bureau of Reference (now superseded by the Standards, Measurement and Testing Programme, SM&T) was applied to the soil samples to extract the metals which are present in exchangeable and acid soluble (i.e. bound to carbonates), reducible (bound to Fe/Mn oxides), and oxidisable forms (bound to organic matter and sulphides) in the soil samples. Trace metals in the soil and vegetable samples were determined using flame atomic absorption spectrometry (FAAS). The total metal contents acquired by summing of metal levels in all the sequential extraction steps were compared with pseudo-total metal levels obtained with aqua regia for all the soil samples. The recovery values obtained by proportioning the results obtained by the BCR procedure to those of the pseudo-total digestion were found to be satisfactory. The limits of detection for the elements investigated were in the range of 0.04 to 0.59 mug mL(-1) for all the extraction stages of the BCR procedure. Similarities among the variables were identified by correlation analysis, principal component analysis and hierarchical cluster analysis. The relationship between the vegetable metal and soil-extractable metal concentrations was examined in order to evaluate the bioavailability of metals.