Gasification of biomass results in release of tar and alkali metal compounds that constitute a significant challenge to the optimization of the gasification process. Here we describe on-line measurements of alkali, condensable tar, and particle concentrations in product gas from a 2-4 MWth dual fluidized bed gasifier, with the aims to characterize typical concentrations and contribute to the understanding of alkali-tar interactions. The influence of bed material, additives, and operational parameters on the concentrations is investigated. Alkali concentrations are measured with a surface ionization detector, and particle and tar concentrations are determined using aerosol measurement techniques. The gasification of wood chips using quartz or olivine as bed material results in an alkali concentration of 30-250 mg m-3, and the observed alkali levels are consistent with a significant release of the fuel alkali content. Addition of ilmenite to a quartz bed and additions of K2SO4 and K2CO3 to an olivine bed influence both alkali and heavy tar concentrations. The additions result in changes in alkali concentration that relaxes to a new steady state in tens of minutes. The concentration of condensable tar is lower for the olivine bed than for the quartz bed, and tends to decrease when potassium or sulfur is added. The concentration of condensable tar compounds is anticorrelated with the alkali concentration when a quartz bed is used, while no clear trend is observed with an olivine bed. An increase in steam flow rate results in a substantial decrease in heavy tar concentration from a quartz sand bed, while the alkali concentration increases slightly with increasing flow rate. This is in contrast to the alkali concentrations observed when using an activated olivine bed, where concentrations are higher and tend to decrease with increasing steam flow rate. The study confirms that several primary methods are available to optimize the alkali and tar behavior in the gasifier, and suggests that on-line monitoring is needed to systematically change the operational conditions and to study the underlying processes.