Equipment and method for accurate and precise concentration, C-14, C-13, and O-18 isotope analysis for CO in background air is presented along with new results for Antarctica and New Zealand. High flow rate cryogenic extraction systems which separate CO after its oxidation to CO2 are used, incorporating a novel ultraefficient cryogenic trap. Air in quantities from a few hundred liters to 2 m3 With CO concentrations from 20 part per billion by volume (ppbv) to 1000 ppbv can be analyzed. The absolute CO concentration is determined volumetrically. The C-13/C-12 and O-18/O-16 ratios are determined by mass spectrometry. For O-18 a correction is applied for the oxygen in the CO2 derived from the oxidant. Carbon 14 is determined by accelerator mass spectrometry. Prior to this the very small CO-derived samples with their high specific activity are diluted accurately. For polluted air the proportional decrease in specific activity with increasing CO levels is confirmed. The (CO)-C-14 abundance and CO concentration in background air in New Zealand and Antarctica are not much different, and both follow a distinct seasonal pattern, in particular (CO)-C-14 which is mainly forced by OH seasonality. The (CO)-C-14 abundance swings between its February minimum of about 6 and its August maximum of about 13 molecules per cm3 air (STP). CO has a smaller seasonality and shows a larger scatter due to local CO sources. The impact of changes in solar activity on (CO)-C-14 for the period considered has been small. Most of the short-term variability in (CO)-C-14 is due to the sampling of different air masses. It appears that interannual OH variations may be reflected in (CO)-C-14 variations. Both C-13/C-12 and O-18/O-16 at Scott Base show large seasonal variation, and the impact of biomass burning and isotopic fractionation in CO destruction are used to try to explain the respective isotopic compositions.