The purpose of the present experiment was to study the ''kindling'' hy
pothesis of alcohol withdrawal stating that exposure to repeated episo
des of alcohol withdrawal results in an increased severity of subseque
nt withdrawal reactions. Two groups of male Wistar rats were subjected
to 13 episodes of 2 days severe alcohol intoxication and 5 days alcoh
ol withdrawal. Animals in the control group (n = 80) developed clinica
l withdrawal signs following each intoxication episode. In the diazepa
m group (n = 80) the withdrawal reactions during episodes 1-9 were blo
cked by intraperitoneal diazepam administration (0-30 mg/kg) 8, 11 and
15 h into withdrawal. During episode 10-13 diazepam treatment was ter
minated and convulsive withdrawal behaviour was observed 9-15 h after
last alcohol dose. The probability of seizure activity during these fo
ur withdrawal episodes was calculated as 0.239 and 0.066 in the contro
l and the diazepam groups, respectively. Based on Monte Carlo simulati
on techniques, this difference was found to be statistically significa
nt (P < 0.05). No differences in the non-convulsive alcohol withdrawal
symptoms tremor hyperactivity and rigidity were detected between cont
rols and diazepam animals after diazepam treatment. It was concluded t
hat the increased convulsive behaviour in the control group was caused
by cumulated kindling-like cerebral alterations during the previous r
epeated alcohol withdrawal phases.