Predicting how both spatial and temporal variation in sea and air temperature influence the distribution of intertidal organisms is a pressing issue. We used data from satellites, weather stations and in situ loggers to test the hypothesis that satellite-derived sea surface temperatures (SSTs) and weather station air temperatures provide accurate estimates of ambient temperature variability on rocky intertidal shores for temporal (hourly for 1 yr) and spatial (10 m to 400 km) variation along the southeast coast of Australia. We also tested whether satellites and weather stations accurately detect the duration, frequency and number of extreme temperature events. Daily mean satellite SSTs and weather station air temperatures were significantly and strongly correlated with intertidal water and air temperatures, respectively (water: r2 = 0.62, air: r2 = 0.63). Nevertheless, depending on location, daily satellite SSTs were up to 6.7°C, and on average 1°C, higher than in situ water temperatures, while daily maximum air temperatures measured by weather stations were up to 23.2°C, and on average 4.2°C, lower than in situ air temperatures. At all locations, the frequency, duration and number of days greater than 30°C, as well as rates of temperature change, were all significantly lower when measured by weather stations. These differences suggest that satellite SSTs and weather stations are ineffective at capturing extremes in intertidal water and air temperature variability. We reinforce the argument that in situ measurements that focus on biologically relevant variation are the only legitimate means of predicting the effects of temperature change on intertidal taxa. © Inter-Research 2011.