A large number of electrochemical concentration cell ozonesondes were calibrated in relation to an ultraviolet photometric absorption ozone instrument prior to using the ozonesondes in atomospheric soundings. The two methods of measuring ozone were in reasonable agreement on the average, but there was considerable variation from one ozonesonde to another. Averaging the individual linear regressions gives 3>ECC=0¿8nbar+(0.96¿0.10)3>UV, where uncertainties represent 90% confidence limits. Applying individual calibration corrections to atmospheric sounding data reduced the mean difference between Dobson spectrophotometric measurements of total ozone overburdens and corresponding values obtained from ozonesonde data from -8.5¿8.2 to -1.4¿7.3% for a series of measurements carried out in 1977. Corresponding uncorrected and corrected differences averaged 3.2¿9.7 and 1.9¿10.6%, respectively, for soundings carried out in 1976.