Abstract Background Diabetes mellitus (DM) is a rapidly growing health concern in Sri Lanka. Diet and physical activity are important modifiable risk factors affecting the incidence, severity and management of DM. The present study aims to evaluate the knowledge and perceptions about dietary patterns and physical activity among a group of adults with DM in Sri Lanka using qualitative research methods. Methods Fifty adults from a cohort of diabetic patients attending the medical clinics at the National Hospital of Sri Lanka were invited for the study. Data were collected via 10 Focus Group Discussions. Verbatim recording and documenting emotional responses were conducted by two independent observers. Directed content analysis of qualitative data was done with the help of NVIVO v10.0. Results Mean age was 61.2 ± 9.9 years and 46 % were males. Mean duration of diabetes was 10.4 ± 7.5 years. All were aware of the importance of diet in the management of DM. But most had difficulty in incorporating this knowledge into their lives mostly due to social circumstances. The majority described a list of ‘good foods’ and ‘bad foods’ for DM. They believed that ‘good’ foods can be consumed at all times, irrespective of quantity and ‘bad’ foods should be completely avoided. Many believed that fruits were bad for diabetes, while vegetables were considered as a healthy food choice. The majority thought that there were ‘special’ foods that help to control blood glucose, the most common being curry leaves and bitter-gourd. Most study participants were aware of the importance of being physical active. However, there was lack of consensus and clarity with regards to type, duration, timing and frequency of physical activity. Conclusions Despite understanding the importance of dietary control and physical activity in the management of diabetes, adherence to practices were poor, mainly due to lack of clarity of information provided. There were many myths with regards to diet, some of which have originated from health care professionals. More evidence is needed to support or refute the claims about ‘special’ foods that the participants believe as being good for diabetes.