On 17 June 1996, Mt Ruapehu in the central North Island of New Zealand erup
ted, creating an ash plume extending for several hundred kilometers over po
pulated areas to the north-east of the mountain. Out study of this natural
experiment aimed to investigate whether the eruption (intervention) had any
adverse effect on the health of those people living within the ashfall are
a, when compared with temporal and spatial controls. The study population c
omprised those people living in areas receiving > 0.25 mm ashfall, and the
control population was unexposed to volcanic ash. Discharge rates from publ
ic hospitals were measured for the 3-month period following the eruption an
d compared with the aggregated corresponding 3-month periods in the previou
s 7 years. The conditions studied were respiratory disease, stroke and isch
aemic heart disease (IHD). Analysis used multivariate Poisson regression mo
dels, with tests for independence and estimation of interaction effects for
change in time before and after the eruption within the study and control
areas. The only detrimental effects detected were a borderline increase in
acute bronchitis (rate ratio, RR = 1.44. p = 0.031) and conflicting evidenc
e for some cardiovascular diseases. If there were any health effects of exp
osure to ash following the eruption, these were too small to be clearly evi
dent with this study design. It is therefore unlikely that any such health
effects will pose a significant public health problem following similar fut
ure eruptions. No targeted public health interventions are therefore indica
ted beyond those that remove people from the immediate area under direct ph
ysical threat from the volcano.