on behalf of the HARVEST Investigators Summary Background and objectives Whether glomerular hyperfiltration is implicated in the development of microalbuminuria in hypertension is not well known. This prospective study investigated the relationship between changes in GFR and microalbuminuria in hypertension.Design, setting, participants, & measurements This study assessed 534 stage 1 hypertensive participants from the Hypertension and Ambulatory Recording Venetia Study (n=386 men) without microalbuminuria at baseline, who were recruited from 1990 to 1995 and followed for a median of 8.5 years. Mean age was 33.968.6 years and mean BP was 146.6610.5/94.065.0 mmHg. Creatinine clearance and 24-hour urinary albumin were measured at study entry and end. Participants were defined as normofilterers (normo) or hyperfilterers (hyper) according to whether GFR was ,150 or $150 ml/min per 1.73 m 2 , respectively. Participants were divided into four groups based on GFR changes from baseline to follow-up end: normo→normo (n=395), normo→hyper (n=31), hyper→hyper (n=61), and hyper→normo (n=47).Results Microalbuminuria progressively increased across the four groups and was 5.3% in normo→normo, 9.7% in normo→hyper, 16.4% in hyper→hyper, and 36.2% in hyper→normo (P,0.001). This association held true in a multivariable logistic regression in which several confounders, ambulatory BP, and other risk factors were taken into account (P,0.001). In particular, hyperfilterers whose GFR decreased to normal at study end had an adjusted odds ratio of 7.8 (95% confidence interval, 3.3-18.2) for development of microalbuminuria compared with participants with normal GFR throughout the study.Conclusions These data support the hypothesis for a parabolic association between GFR and urinary albumin in the early stage of hypertension.