A sensitive and rugged residue analysis method was validated for the estimation of dithiocarbamate fungicides in a variety of fruit and vegetable matrices. The sample preparation method involved reaction of dithiocarbamates with Tin(II) chloride in aqueous HCl. The CS 2 produced was absorbed into an isooctane layer and estimated by GC–MS selected ion monitoring. Limit of quantification (LOQ) was ⩽40μgkg −1 for grape, green chilli, tomato, potato, brinjal, pineapple and chayote and the recoveries were within 75–104% (RSD<15% at LOQ). The method could be satisfactorily applied for analysis of real world samples. Dissipation of mancozeb, the most-used dithiocarbamate fungicide, in field followed first+first order kinetics with pre-harvest intervals of 2 and 4days in brinjal, 7 and 10days in grapes and 0day in chilli at single and double dose of agricultural applications. Cooking practices were effective for removal of mancozeb residues from vegetables.