A survey was conducted between 1995 and 1997 to assess the impact of introduction of unleaded petrol and other public health measures on the blood lead level of the population. The geometric mean blood lead level of 269 government employees as determined by graphite furnace atomic absorption spectroscopy, was 66.0 microg/l, much lower than that recorded before introduction of lead-free petrol. Using multiple regression analysis, factors significantly associated with blood lead levels were: exposure to traffic, age (>50 years) and active smoking. Passive smoking, exposure to recent paint work, consumption of alcohol and traditional medicine were found not to be significantly associated with the blood lead level.