Late Quaternary branched GDGT dataset from Lake Baikal, Russia, and an expanded calibration dataset of globally distributed lacustrine brGDGT surficial sediment samples. Also included are a new dry bulk density composite section and orbitally tuned age model for cores 340-PC1, 333-PC2, and 340-TC1, for which the brGDGT data are included.We measured new brGDGT samples from Lake Baikal and compiled previously published brGDGT distributions from lake surface sediments and their catchment soils. For our new samples from Lake Baikal, lipids were extracted by an accelerated solvent extractor (ASE) 350 using 2:1 dichloromethane: methanol. Total lipid extracts were spiked with 100 µL of a general recovery standard (20.5 ng/µL 5α-androstane, 20.5 ng/µL steryl stearate, 20.5 ng/µL 1,1'-binapthyl, 20.5 ng/µL cis-11-eicosenoic acid, 20.5 ng/µL 19-methyleicosenoic acid, 20.5 ng/µL C20:1 Δ11-eicosenol, and 20.5 ng/µL 5α-androstan-3β-ol), desulfurized by sequential addition of activated copper wire, and evaporated under N2 prior to elution on a dry-packed LC-NH2 column to separate neutral (4 mL 2:1 dichloromethane: isopropanol), acid (4 mL 4% acetic acid in diethyl ether), and polar (4 mL methanol) fractions. Neutral and polar fractions were recombined and again dried under N2 prior to elution on a wet-packed silica gel (0.5 g, 60 Å, 70–230 mesh, Millipore) column with 3 mL hexanes (apolar / aliphatic), 4 mL dichloromethane (semi-polar / aromatic), and 4 mL methanol (polar / alcohols).The polar/alcohols (methanol) was further chromatographed over alumina oxide (0.85 g, J.T. Baker, 0537-01) by elution of 4 mL 9:1 hexane:dichloromethane, 4 mL 1:1 dichloromethane: methanol (containing GDGTs), and 4 mL 100% methanol. The 1:1 dichloromethane: methanol fraction was then dried over N2 and shipped to Brown University, where they were analyzed for GDGTs on an Agilent/Hewlett Packard 1100 series liquid chromatograph mass spectrometer (LC-MS) using two UHPLC columns (BEH HILIC columns, 2.1 x 150 mm, 1.7 μm, Waters) in series using a modified version of the protocol of Hopmans et al. (2016). Prior to analysis, samples were passed through a 0.45 μm filter and spiked with a known quantity of a C46 diol internal standard.Selective ion monitoring was used to measure m/z 1302, 1300, 1298, 1296, 1292, 1050, 1048, 1046, 1036, 1034, 1032, 1022, 1020, 1018, and 744. Peak areas were quantified manually.Cores 333-PC2, 340-TC1, and 340-PC1 were correlated to generate a continuous composite section based on dry bulk density. Our 333-340 composite section was then assigned age by correlating dry bulk density to the orbitally-tuned BDP-96 biogenic silica record (Prokopenko et al., 2006), supplemented by published radiocarbon dates (Peck et al., 1994) and an assumed core top age of 0 ka.