The mean annual cycle of elemental carbon (EC) has been deduced from a 7-year series of measurements made at Cape Grim, Tasmania. The most conspicuous features are a spring (October) maximum and summer (January) minimum. These are also found in the mean annual cycle at Cape Grim of carbon monoxide thought to be produced at the same time and in the same regions (predominantly much further north). Differences in detail in the cycles are consistent with the different removal mechanisms involved. However, the very considerable similarities between the two curves suggest that seasonal variations on OH radical concentration is not the main cause of CO's seasonal cycle and also that seasonal variations in EC scavenging is not the main cause of EC's seasonal cycle. The seasonal variation of radon, also of continental origin, is out of phase with the variations of EC and CO, showing that strength of poleward transport is not the cause of the seasonal change of EC and CO. Methane, only a small portion of which is produced by biomass burning, has a seasonal variability closely similar to that of both EC and CO suggesting that biomass burning is the only source with significant seasonality of EC, CO, and CH4. The broad summer maximum of mean 7Be concentrations shows that the spring maximum of EC and CO is not due to upper transport with enhanced spring mixing. It seems unlikely that interhemispheric transport plays a major role in the seasonal cycle of EC, CO and CH4 because of the phase lag between the predominantly northern hemisphere produced tracers and that of the three first named tracers. Comparison of the tracer cycles at Cape Grim, Cape Point, S. Africa and over Antarctica suggests that the spring maximum could be due to a peak in production of combustion products from biomass burning at low latitudes which is propagating throughout the southern hemisphere. DOI: 10.1034/j.1600-0889.1990.t01-3-00005.x