The public health benefit of conjugate pneumococcal vaccine (PCV) in preventing pneumonia would only be appreciated if the tool used for measuring "pneumococcal pneumonia" had good sensitivity. Exploratory studies in South Africa indicate that the sensitivity of "radiologically-confirmed pneumonia" (CXR-AC) underestimates the burden of pneumococcal pneumonia prevented by PCV by as much as 63%. The use of alternate markers such as C-reactive protein enhance the ability of measuring the burden of pneumonia preventable by PCV. A broadened definition of "pneumococcal pneumonia" which includes episodes of pneumonia associated with CXR-AC and those associated with an abnormal chest radiograph other than CXR-AC associated with a CRP of > or =40mg/l should be considered an a priori outcome in future PCV efficacy trials.