Background In 2011, the IFCC Committee on Reference Intervals and Decision Limits (C-RIDL) initiated a worldwide multicenter study on references values facilitating the implementation of country-specific reference intervals (RIs). There has been no well-designed RI study in India. This study aims to derive RIs for 33 major biochemical analytes in carefully selected healthy Indians as defined in C-RIDL protocol. Methods A total of 512 healthy Indians were recruited. Sera collected from overnight fasting blood samples were measured collectively for the analytes. Multiple regression analysis (MRA) and nested analysis of variance (ANOVA) were used to identify the potential sources of variation (SV) of test results. RI were derived by both parametric and non-parametric methods for comparison. The need for secondary exclusion by latent abnormal values exclusion (LAVE) method was examined. Results MRA results indicated that both age and BMI were apparent SV for many analytes in both sexes. ANOVA revealed that partition of RIs by gender and age was required for 17 analytes (TC, HDL-C, TG, hsCRP, ALB, AST, ALT, ALP, GGT, TBil, Urea, CRE, UA, Fe, TTR, CK and IgM) and 5 (Glu, ALB, TC, ALP and Urea), respectively. RIs by parametric method were generally narrower than by non-parametric method, reflecting distorted peripheral distributions of test results. The LAVE method had no appreciable effect on RIs possibly due to inconsistency among abnormal values of related analytes. Conclusions This study has for the first time provided comprehensive RIs information in healthy Indians. The final RIs adopted were those derived by parametric method without LAVE procedure.