One of the most difficult problems of behavioral research on human sexualit
y and drug use is its reliance on people's self-reports about their behavio
r: Given the essentially private nature of sexual and drug-taking practices
, many aspects of these practices are difficult even impossible, to validat
e. The lack of a gold standard for self-reports is particularly troubling w
ith regard to HN-related behavior To the extent that self-reports are affec
ted by response bias, the prevalence of high risk practices will be underes
timated hindering efforts to slow the spread of HIV In this paper the exten
t of response bins in seven measures of high-risk injection-related and sex
ual behaviors is estimated Self-reported high-risk injection practices may
be significantly to substantially underreported due to response bias, where
as self-reports of some high-risk sexual practices do not appear to be cont
aminated by response bias. In regression analyses, the response bias in sel
f-reports of needle sharing and other high-risk injection practices was mod
eled, and estimates of the probable actual prevalence of these practices de
rived. In situations in which behavioral self-reports cannot be validated s
uch derivations may help in arriving at more accurate estimates of behavior
.