In this systematic review, walking (a generally accessible activity for a largely sedentary population) was assessed as a preventive risk factor for development of fatal and non-fatal cardiovascular disease (CVD). PubMed, CINHAL and reference list searches identified 21 peer-reviewed publications examining walking in relation to CVD; studies assessing active transportation were excluded. Generally, there were dose-dependent reductions in CVD risk with higher walking duration, distance, energy expenditure and pace. Associations appeared to be stronger for ischaemic stroke than other CVD outcomes such as coronary heart disease or haemorrhagic stroke. Adjustment for clinical CVD risk factors, obesity or other types of physical activity generally attenuated but did not eliminate associations. Because functional status may be an important determinant of walking behaviour in adults, potential bias due to pre-existing illness is of concern in all studies reviewed, particularly in case–control studies which ascertain walking retrospectively and yielded the strongest associations. Study findings were consistent with current physical activity recommendations, but there is a need for improvements in measurement of walking and other CVD risk factors, more thorough control for pre-existing illness, examination of mediating or moderating conditions such as obesity and other analytical issues.