Abstract. I present an analysis of ozonesonde data with a focus on using these data to evaluate stratospheric models of transport and chemistry. Ozonesondes are the only instruments that provide year-round profiles of ozone throughout the lowermost stratosphere for middle and high latitudes of the Northern Hemisphere. I show vertical profiles and the annual cycle of ozone at selected pressure levels and stations and use TOMS data to evaluate the spatial bias of the sonde stations with respect to zonal mean values of column ozone. Between 10 and 30% of the ozone column is located between 100 hPa and the tropopause at middle and high latitudes, and this region drives much of the seasonal variation of the ozone column. The sonde data allow quantification of the buildup of ozone in the lowermost stratosphere of the Northern Hemisphere in winter and of its loss in late spring and summer. The amount of ozone between the tropopause and 100 hPA in the Northern Hemisphere decreases from 175 to 75 Tg from March to September, with the maximum rate of decrease in May to June; about half of the decrease is caused by the increase in the height of the tropopause. There is a lag of about a month in the time when ozone starts to accumulate in the lowermost stratosphere, from September near 40øN to October or November near 80øN, and a similar lag in the time when ozone starts to decrease, from February or March near 40øN to March or April near 80øN. I propose that the sonde data be used to provide a test of transport in two-and three-dimensional models in addition to those provided by long-lived tracers, such as N20, CFCs, CO2, CH4, and their correlation patterns, and age of air. The annual variation of the ozone content of the lowermost stratosphere in the Northern Hemisphere is suggested as a particularly useful constraint on transport parameterizations. I make recommendations for the use of particular stations and provide summary statistics for ozone concentrations and for the mean tropopause height for each station.