The effective evaluation of preventive activities depends on the ident
ification of indicators and the selection of appropriate outcome measu
res which reflect the goals of the intervention. An increase in condom
use has been seen as a positive sign of the impact of HIV/AIDS public
education. This paper examines possible sources of data relating to c
ondom use in the contest of assessing public response to the AIDS epid
emic, with particular reference to methodological challenges presented
by each; issues relating to the validity of data, problems of interpr
etation and the scope for improvement. A multiple indicator approach,
using several types of data in unison, is advocated. Conclusions drawn
from the multiple indicator approach are likely to be firmer and soun
der than those drawn from the single indicator approach, and are more
likely to offer insight into the mechanisms which influence particular
outcomes.