Abstract. Boreal forest fires are currently recognized as a significant factor in climate change and air quality problems. Although emissions of biomass burning products are widely measured in many regions, there is still lack of information on the composition of wildfire emissions in Siberia, a region known for its severe wildfire activity. Emission ratios (ERs) are important characteristics of wildfire emissions as they may be used to calculate the mass of species emitted into the atmosphere due to combustion of a known mass of biomass fuel. We analyze observations of carbon dioxide (CO 2 ), carbon monoxide (CO), methane (CH 4 ), total nonmethane hydrocarbons (NMHCs), nitrogen oxides NO x (= NO + NO 2 ), particulate matter (PM 3 ), and black carbon (BC) within two forest fire plume transects made by the moving railway observatory during TRanscontinental Observations Into the Chemistry of the Atmosphere (TROICA) expeditions. Slopes in linear regressions of excess levels of the pollutants are used to obtain ER CO / CO 2 = 10-15 %, ER CH 4 / CO = 8-10 %, ER NMHC / CO = 0.11-0.21 % ppmC ppmC −1 , ER NO x / CO = 1.5-3.0 ppb ppm −1 , ER PM 3 / CO = 320-385 ng m −3 µg m −3 −1 , and ER BC / CO = 6.1-6.3 µg m −3 ppm −1 , which fall within the range of uncertainty of the previous estimates, being at the higher edge for ER CH 4 / CO , ER NMHC / CO , and ER PM 3 / CO and at the lower edge for ER NO x / CO . The relative uncertainties comprise 5-15 % of the estimated ER CH 4 / CO , ER NMHC / CO , and ER PM / CO and 10-20 % of ER NO x / CO , ER CO / CO 2 , and ER BC / CO . The uncertainties are lower than in many other similar studies and associated mainly with natural variability of wildfire emissions.