The Darlington/Northallerton prospective study of asthmatics referred to secondary care started in 1983, with review and new entry at 5-yr intervals. The principal outcome measures are: mortality (presented here), best function and therapeutic step. All adult asthmatics with > or = 15% peak flow (PEF) reversibility to > or = 200 l min-1 were included. Socio-demographic variables, PEF and spirometry were recorded prospectively. Best vital capacity (FVC) and PEF were assessed according to protocol. The mortality of the original cohort after 10 yr was expressed as standardized mortality ratio (SMR) against the local population, with history and pulmonary function at entry as explanatory variables. Ninety-five per cent follow-up was achieved in 628 subjects, with 173 deaths (29.1% of those traced). The excess death rate was nearly 50% (SMR 1.47, 95% CI 1.26-1.71), with 56% of deaths due to respiratory disease (expected 10%). After allowance for age and sex, there was a consistent inverse relationship between mortality and entry best FVC, increased risk of death 1.51 (95% CI 1.33-1.72) per 10% deficit of best FVC predicted. The risk of respiratory death was eight times greater, and of non-respiratory death three times greater, in the lowest compared with the highest quartile of best FVC. There were no interactions with smoking, but possible enhancement of the effect in the socially deprived. Best FVC was a particularly powerful predictor of mortality in subjects < 65 years at entry, in whom 64% of the excess deaths occurred. Most of the excess in respiratory deaths was not due to acute severe asthma but to the development of chronic obstructive pulmonary disease (COPD), as defined functionally, irrespective of smoking habit which made no further contribution to mortality.