Obesity is increasing in severity and prevalence representing a major public health issue within the UK. Studies suggest that, at any one time, most US and UK adults are trying to lose or maintain weight ( Offer, 2001). As a result the demand for weight loss schemes is high. As obesity is most prevalent amongst low income groups ( James et al., 1997 ), the financial implication of following such programmes requires investigation. This study aimed to quantify and compare the costs incurred by subjects undertaking one of four commercially available weight loss programmes. Subjects comprised of the Surrey cohort (n = 35) of a multi-centre weight loss trial and were randomized to one of four programmes: Weight Watchers (WW), the Slim-Fast Plan, Rosemary Conley's Diet and Fitness Plan (RC) or the Dr Atkins’ New Diet Revolution (Atkins). Participants were all aged between 21 and 60 years, had an initial body mass index in the range of 27–36 kg m−2 and were otherwise healthy. Seven-day estimated food diaries were completed at baseline and after 2 months of dieting and were used to calculate the weight of food consumed. The cost was subsequently based on the prices quoted on the Tesco's website. Total food and alcohol were calculated for each subject at baseline and at 2 months and total programme cost was calculated at 2 months (food costs and any additional costs as a result of programme membership fees, supplementary vitamins, etc). Food group expenditure was calculated to investigate the contribution of meat, fruit, vegetables and alcohol to total costs and change in costs. Parametric and nonparametric statistics (according to the data distribution) were used to compare change in mean costs over time (paired t-tests, Wilcoxon signed ranks test) and by programme followed (anova). No significant differences in baseline costs were observed between the four groups with whole sample mean food costs of £20.53 (SD 5.10). The Atkins group showed a significant increase in food expenditure at 2 months (Z = −2.24, p = 0.03) whereas all other groups decreased their spending on food, significantly so for the RC and WW subjects (Z = −1.96, p = 0.05; Z = −2.67, p = 0.01). As a result, significant differences in percentage change in food expenditure were recorded across groups (2 = 16.47, p = 0.001) with the Atkins group significantly different from both RC and WW groups. Although all programmes incurred initial ‘start-up’ costs, only the Atkins diet had ongoing total programme costs at 2 months that were significantly greater than baseline expenditure. Increases in meat expenditure were significantly greater within the Atkins group compared with subjects following the other programmes (F = 4.01, p = 0.02). However, contrary to expectations, no significant increases in spending on fruit and vegetables were seen, with a significant decrease in expenditure on fruit observed across the whole sample and within the WW group (Z = −2.40, p = 0.02 and Z = −2.07, p = 0.04 respectively). This study suggests that commercial weight loss programmes, including those requiring the purchase of meal replacements, are not necessarily associated with an increase in food expenditure. However, the Atkins diet did result in an increase in food expenditure and as all the programmes are subject to potentially prohibitive ‘start-up’ costs, such as joining fees, the issue of access for those population groups most at risk from obesity remains.