This paper presents a critical review of the role of froth recovery in laboratory flotation kinetics tests. By conducting tests in the standard lab equipment (the Denver cell), it is demonstrated that, for typical scraping rates, the froth recovery is significantly lower than the 100% that is commonly assumed when interpreting lab kinetics data. Furthermore, it is shown that the curve of overall rate constant versus froth residence time, as defined by the scraping rate, is not linear, but increases quickly at faster scraping rates. These findings have important implications for scale-up. For one, differences in froth recoveries at the lab scale can lead to significant error in the modeled plant recoveries. For two, they undermine a key assumption used to derive and validate the linear relationship between collection rate constant and bubble surface area flux. This casts doubt on the assertion that it is only the collection rate-rather than the froth recovery, interface recovery, or some combination thereof-that is responsible for the observed collinearity between the bubble surface area flux and the overall rate constant. (C) 2014 Elsevier Ltd. All rights reserved.