This study was designed to evaluate the accuracy and feasibility of use of three commercially available portable blood glucose meters to measure amniotic fluid glucose(AFG) levels as compared to an accepted laboratory standard. A prospective study of amniotic fluid from 101 consecutive amniocenteses was performed. Glucose concentration in the amniotic fluid was assessed by hexokinase method in our hospital laboratory (control) and by using three portable meters: Advantage (ADV) (Boehringer Mannheim), Glucometer Elite (ELT) (Bayer), and One Touch II Hospital (T-2) (Lifescan). Twenty consecutive amniotic fluid samples were sent to the laboratory in two vials, the first without additive and the second with potassium oxalate to prevent metabolic activity, to assess the effect of cellular metabolism and time delay on amniotic fluid glucose concentrations. Data are reported as mean ฯฎSE and were assessed by one-way ANOVA. Of the 101 patients studied, 29 were of gestational age ี20 wks. The remaining 72 patients were ฯฝ20 wks. All three ambulatory meters demonstrated a linear relationship with control (all P ฯฝ 0.001). Given a slope of almost 1.0 (m ฯญ 0.94) and a y-intercept approaching zero (b ฯญ 4.3), the OT2 proved to correlate best with control. ELT: (r 2 ฯญ 0.55, m ฯญ 0.79, b ฯญ 22.2) and ADV: (r 2 ฯญ 0.74, m ฯญ 1.45, b ฯญ 16.9) both overestimated amniotic fluid glucose. When AFG was ฯฝ30 mg/dl via laboratory standard, OT2: (r 2 ฯญ 0.78, m ฯญ 1.05, and b ฯญ ฯช2.20, P ฯฝ 0.001), ADV: (m ฯญ 1.02, b ฯญ 24.1, r 2 ฯญ 0.12, P ฯญ 0.133). The One Touch II Hospital accurately predicted amniotic fluid glucose at the bedside with excellent correlation including with laboratory standard glucose levels ฯฝ30 mg/dl. ADV and ELT proved too inaccurate for clinical use. Control samples were not affected by additives or time delay. These findings confirm that AFG determinations can be obtained rapidly with the OT2 meter at the bedside.