Rationale: In animal models of drug self-administration, response rates oft
en decrease with dose suggesting that a regulative process may mask the rei
nforcing effects of the drug. Objective: The purpose of the present experim
ents was to dissociate the role of regulative and reinforcement processes i
n intravenous cocaine self-administration by rats using a paradigm that exp
licitly distinguishes between drug-seeking and drug-taking. Methods: Rats w
ere trained to respond for intravenous cocaine (0.25 mg/infusion) under a h
eterogeneous chain (tandem FR1 RI 30 s) FR1 schedule of reinforcement using
different levers in the first (seeking) and second (taking) links of the c
hain. After 10 days of training, rats were switched to one of three doses o
f cocaine (0.08, 0.25, or 0.5 mg/infusion) and self-administration patterns
were recorded for a further ten sessions in experiment 1. In experiment 2,
a time-out (TO) period (0, 4, or 12 min) was imposed between successive cy
cles of the chain schedule. Finally, the effect of allowing animals to per
form a drug-taking response on subsequent drug-seeking was assessed in expe
riment 3. Results: Having verified that seeking responses for a conventiona
l reinforcer (sucrose) were sensitive to changes in reward magnitude, exper
iment 1 demonstrated that the number of self-administered infusions was inv
ersely related to dose whereas the latency to initiate drug-seeking increas
ed with dose. Variations in the cocaine dose had no reliable effect on the
number of drug seeking response per cycle of the chain schedule. The effect
of dose on the latency to initiate drug-seeking was reversed in experiment
2 with increasing TO periods. Moreover, at the longest TO period, drug-see
king responses per cycle increased and the latency to initiate drug seeking
decreased with dose. Experiment 3 showed that the latency to drug-seek for
the low dose was reduced dramatically when the first drug-seeking response
was preceded by a drug-taking response, even when this response did not pr
oduce a drug infusion. Conclusions: The overall pattern of results suggests
that drug-seeking and drug-taking are controlled by three interacting proc
esses: a regulative process depresses drug-seeking in the short-term; behav
ioral activation enhances drug-seeking and is sustained over longer interva
ls by higher drug doses; the reinforcing effect of cocaine increases with d
ose once the satiety producing effects of the drug dissipate.