A major question concerning the observed long-term changes of zonal average tow ozone has always been that the spatially limited ground-based ozone sampling sites are susceptible to a sampling problem. That is, the regional (or station) averages are influenced by shifts of the ozone wave patterns with respect to the sites such that a trend may be indicated that is not necessarily indicative of the actual zonal average. In order to help determine whether these sampling errors are short-term random features or have long-term components, we have utilized available synoptic analyses of the 100 mb height patterns (1964–16) and the observation that the 100 mb heights and tow 03 patterns in the midlatitudes of the Northern Hemisphere are negatively correlated. Accordingly, the ridge-trough patterns in the 100 mb height field at 50°N from 1964–76 are sampled over the domain of the ground-based 03 observing sites and a zonal average of 100 mb heights calculated using, the area-weighting functions of Angell and Korshover. These zonal averages are compared with the actual zonal average computed from all data and the results noted as a function of time. Utilizing linear 100 mb height– total 03 regression relationships, the zonal average total ozone sampling error is on the order of ±2% for midlatitudes of the Northern Hemisphere with a long-term component. With this result, the general shape of the midlatitude O3 trends determined from the ground-based observations appears to be real and not an artifact of the spatially limited ground-based sample. In fact, the increase of ozone from the mid-1960' to early 1970's may be even greater than previously suggested.