Objective: We analyzed data on all drug-treatment admissions in San Fr
ancisco County over a 4-year period (n = 35 460) to evaluate the poten
tial negative effects of the San Francisco needle-exchange program. Me
thods: Admissions in the 2 years preceding implementation of the excha
nge program (1987-1988) were compared with admissions in the 2 years f
ollowing implementation (1989-1990). Results: No negative consequences
of needle exchange were detected. Specifically, the presence of the e
xchange program was not associated with (1) increases in injection dru
g use, (2) increases in needle-sharing behavior, or (3) changing drug-
use behavior from non-injection to injection. We also compared high-dr
ug-use neighborhoods with and without local needle-exchange sites. Nei
ghborhoods without needle-exchange sites showed a greater increase in
proportion of admissions for injection drug use, and in frequency of i
njection, over time. Conclusions: Because of methodological limitation
s, our findings are preliminary. Longitudinal studies comparing needle
-exchange and non-exchange cohorts are needed to further evaluate effe
cts of this intervention.