Objective To estimate the gestational age-specific serum creatinine (sCr) in pregnancy. Design A population-based prospective cohort study. Settings Anuradhapura district, Sri Lanka Population or sample Study group: Healthy pregnant women with a period of gestation less than 12 weeks and without pre-existing medical conditions. Comparison group: A sample of non-pregnant reproductive age females from a population-based renal screening. Methods Baseline data were compared among pregnant and non-pregnant groups and the pregnant women were followed-up until the end of the second trimester. Main Outcome Measures Gestational period-specific sCr. Results A total of 2,259 pregnant women and 2.012 non-pregnant women were recruited. The mean (SD) sCr of the 2,012 nonpregnant women was 62.8(12.4) μmol/L, with the 97.5th percentile of 89.0 μmol/L. The mean (SD) eGFR was 105.1(27.9) mL/min/1.73 m2. At 4-7, 8-9, 10-12, 24-27 and 28-30 weeks of pregnancy, the mean sCr was 55.1, 52.7, 51.0, 47.2, and 49.3, while the 97.5th percentile for sCr was 72.4, 69.2, 69.3, 63.9, and 66.0 μmol/L, respectively, in the sample of pregnant women. In the first and second trimesters, the average sCr value was 84.7% and 76.4% of that of the nonpregnant group, respectively. The mean eGFR increased up to 129.4 mL/min/1.73 m2 in the 24th week of gestation. The analysis of cohort data clearly confirmed a significant reduction in sCr with advancing pregnancy (p<0 .001). Conclusions This study confirms that the precise normative data needs to be considered in the interpretation of sCr in pregnancy, based on the period of gestation.