Changes with time in the anti-factor Xa activity of several heparins were determined in freshly autoclaved and unautoclaved dextrose solutions. In the former, activity was raised, in the latter a reversible fall occurred at certain heparin concentrations, and the term "dextrose effect' is applied to the difference in activity in the two types of dextrose solutions. A simple dependence on heparin concentration for the rise in activity in autoclaved dextrose solutions contrasted with a threshold heparin concentration in the unautoclaved dextrose solutions. Above this a fall and below it a rise in activity were demonstrated with one factor Xa method. The pH of the two dextrose solutions differed by one unit but pH was not the principal factor in the effect; salt tended to eliminate the dextrose effect which was found mostly with the high and to some extent also with the low molecular weight fractions of pharmacopoeial heparin. There was no evidence for heparin degradation but activity changes, which differed for different heparins, are thought to be associated with the effects of heparin-dextrose interaction modified in the case of autoclaved dextrose by by-products of autoclaving. Generally, proprietary dextrose solutions gave variable results.