Rationale and objectives: Previous studies have shown that insomniacs self-
administer hypnotics at high nightly rates. This study determined whether p
rior experience with different treatment regimens (i.e., instructions and c
apsule availability) would alter the previously observed high hypnotic self
-administration rates. Methods: Sixty-four healthy men and women with (n=32
) and without (n=32) insomnia, 21-55 years, self administered placebo or tr
iazolam (0.25 mg) after different prior treat ment regimens. They received
one of three different treatment regimens enforced for 11 nights: a capsule
each night, a capsule as needed, or a capsule every third night. On 14 sub
sequent nights they choose to self administer a capsule or not, placebo dur
ing 1 week and triazolam (0.25 mg) the other (counterbalanced in order). Re
sults: Insomniacs self-administered more capsules than normals and triazola
m was self-administered more than placebo. For both groups, treatment regim
en had a minimal effect on capsule self-administration. During the treatmen
t phase, triazolam improved self-ratings of sleep relative to placebo. Duri
ng the choice phase, nightly variations in self-rated sleep predicted self-
administration of a capsule on the following night, regardless of whether t
he capsule was active drug or placebo. Conclusions: The data of this study
are consistent with the view that hypnotic self-administration by insomniac
s is therapy-seeking behavior and not drug abuse.