We studied the role of cold waves and heat waves on major power outages in the metropolitan area of Buenos Aires. Impacts of events occurring in the tails of distributions were assessed estimating deviations of minimum temperature, maximum temperature and hourly electricity consumption with respect to statistically derived thresholds and using three sets of data: temperature observations (1911-2013); major power outages reported in a disaster database (1971-2013) and hourly electricity consumption (2006-2013). These deviations (exceedances) proved to be adequate indicators of the stress posed by extreme temperature events to the electricity distribution system leading to major blackouts. Based on these indicators, we found that the electricity distribution system was under similar stress during cold waves or heat waves, but it was much more vulnerable to heat waves (three blackouts under cold waves against 20 under heat waves between 2006 and 2013). For heat waves, the results of a binomial regression logistic model provided an adequate description of the probability of disastrous supply interruptions in terms of exceedances in extreme temperatures and electricity consumption stress. This approach may be of use for other cities wishing to evaluate the effects of extreme temperature events on the electricity distribution infrastructure.