Within- and between-population variation in spirometric measurements has been addressed in a systematic review of spirometric measurements from 29 studies published between 1965 and 1990 involving 9,690 men and 2,638 women of sub-Saharan African ancestry; FVC and FEV1 were age- and height-standardized at BTPS. Between- population differences were related to geographic region and sample source (workforce versus community). The effects of altitude, variation in sample mean height, and year of study publication were also significant variables in multivariate models explaining between-population differences. Altitude was the most important variable with an effect of 263 ml/1,000 m (95% confidence interval [CI] 120-410) on FVC in men. In women, variation in sample mean height was also important, with taller populations having larger values for FVC (64 ml/cm). In men, secular trends were evident, with date of study associated with negative trends among men in the U.S.A. (-35.3 ml/year), and positive trends (14.7 ml/year) in other regions. These differences in trend could not be explained. Population selection factors, altitude, date of study, and other biological sources of variation need to be taken into account in evaluating between- and within-population comparisons of spirometric measurements.