OBJECTIVE: To determine whether a cash-for-work programme during the annual food insecurity period in Bangladesh improved nutritional status in poor rural women and children. METHODS: The panel study involved a random sample of 895 households from over 50,000 enrolled in a cash-for-work programme between September and December 2007 and 921 similar control households. The height, weight and mid-upper arm circumference of one woman and child aged less than 5 years from each household were measured at baseline and at the end of the study (mean time: 10 weeks). Women reported 7-day household food expenditure and consumption on both occasions. Changes in parameters were compared between the two groups. FINDINGS: At baseline, no significant difference existed between the groups. By the study end, the difference in mean mid-upper arm circumference between women in the intervention and control groups had widened by 2.29 mm and the difference in mean weight, by 0.88 kg. Among children, the difference in means between the two groups had also widened in favour of the intervention group for: height (0.08 cm; P<0.05), weight (0.22 kg; P<0.001), mid-upper arm circumference (1.41 mm; P<0.001) and z-scores for height-for-age (0.02; P<0.001), weight-for-age (0.17; P<0.001), weight-for-height (0.23; P<0.001) and mid-upper arm circumference (0.12; P<0.001). Intervention households spent more on food and consumed more protein-rich food at the end of the study. CONCLUSION: The cash-for-work programme led to greater household food expenditure and consumption and women's and children's nutritional status improved.