Inorganic carbonates are often used to establish radiocarbon (C-14) chronologies for lake sediments when terrestrial plant remains (TPR) are rare or when bulk organic matter is insufficient for dating, a problem that is common for many lakes in arid regions. However, the reservoir effect (RE), as well as old carbon contributed from the lakes catchment make it difficult to establish reliable chronologies. Here we present a systematic study of inorganic C-14 ages of two lake-sediment sequences, one from a small-enclosed saline lake - Lake Gahai in Qaidam Basin, and the other from a large freshwater lake - Lake Bosten in Xinjiang. Modern dissolved inorganic carbon (DIC) of the lakes, paleo-lake sediments exposed in the catchment, and mollusk shells in core sediments from Lake Gahai were dated to assess the RE and the contribution of pre-aged carbon to the old ages in the cores. We propose a statistical regression to assess more than one RE for the C-14 carbonate ages within our sedimentary sequences. Old radiocarbon ages contributed by detrital carbonates were assessed by comparing the ages of mollusk shells with those of carbonates at the same sediment depths. We established the RE of the authigenic component and assessed detrital old carbon contributions to our two sites, and this was used to correct the C-14 ages. Based on this approach, we developed age models for both cores, and tested them using Pb-210 ages in both cores and TPR-based C-14-ages recovered from Lake Bosten. We further tested our age models by comparing carbonate-based oxygen isotope (delta O-18) records from both lakes to an independently-dated regional speleothem delta O-18 record. Our results suggest if sedimentary sequences are densely dated and the RE and the contribution of old carbon from detrital carbonates can be ascertained, robust chronological frameworks based on carbonate-based C-14 determinations can be established. (C) 2016 Elsevier Ltd. All rights reserved.