Flat coil-coated galvanised steel samples from two sources were cut, bent and subjected to exposure at two natural weathering sites for five years. Results were compared to two accelerated salt-spray replicated tests. The paper discusses methods of measuring corrosion and compares performance of coated products. Cyclic weathering followed by salt-spray testing produced similar results to natural exposure tests when substrate pitting was examined but not when coating blistering was examined. Salt-spray pitting and blistering results were not similar to natural exposure results. Pitting of the substrate was regarded as a more accurate measure of coated metal performance than coating blistering, though removal of the coating is necessary. There was a large variation in T bend results for all exposures. Conclusions were that combination test methods should be further examined. Cyclic weathering followed by salt-spray testing proved a useful indicator of natural exposure results, but was less useful for product comparison work.