Rats with chronically implanted jugular catheters and cortical EEG ele
ctrodes were allowed 24 hr per day access to morphine infusions (30 mu
g/kg) contingent on lever-pressing. Yoked control subjects received th
e same number and pattern of infusions as contingent subjects in an ad
jacent cage. Six subjects were studied, with data analyzed over 84 day
s of contingent and 63 days of yoked control (noncontingent) treatment
. Self-administration rates were positively correlated with previous e
xposure to morphine. EEG total power was reduced in both contingent an
d yoked subjects during periods of self-administration. Averaging of t
rends in EEG power over time across multiple episodes of self-administ
ration showed a greater reduction in power for contingent subjects tha
t preceded the majority of morphine infusions. Desynchronization (dimi
nished EEG amplitude) for contingent subjects at the onset of self-adm
inistration is probably related to lever-pressing activity. Ultradian
describes biological rhythms in which the duration of one cycle (perio
d) ranges from several minutes to values less than 24 hr. In both cont
ingent and yoked subjects at intermediate or high levels of self-admin
istration, increases occurred in the period and amplitude of ultradian
cycles in EEG total power. Responsiveness to the light-dark cycle was
also diminished at these levels of self-administration. In conclusion
, morphine self-administration at intermediate or high levels disrupts
both diurnal and ultradian rhythms in EEG total power for both contin
gent and yoked subjects. The ultradian EEG pattern associated with gre
ater levels of morphine self-administration resembles diminished varia
tion in EEG power that occurs normally during the inactive rest phase
of the diurnal cycle.