BACKGROUND AND AIMS: For the last decade, there has been an increasing global interest in using biochar to mitigate climate change by storing carbon in soil. However, there is a lack of detailed knowledge on the impact of biochar on the crop productivity in different agricultural systems. The objective of this study was to quantify the effect of biochar soil amendment (BSA) on crop productivity and to analyze the dependence of responses on experimental conditions. METHODS: A weighted meta-analysis was conducted based on data from 103 studies published up to April, 2013. The effect of BSA on crop productivity was quantified by characterizing experimental conditions. RESULTS: In the published experiments, with biochar amendment rates generally <30 t ha⁻¹, BSA increased crop productivity by 11.0 % on average, while the responses varied with experimental conditions. Greater responses were found in pot experiments than in field, in acid than in neutral soils, in sandy textured than in loam and silt soils. Crop response in field experiments was greater for dry land crops (10.6 % on average) than for paddy rice (5.6 % on average). This result, associated with the higher response in acid and sandy textured soils, suggests both a liming and an aggregating/moistening effect of BSA. CONCLUSIONS: The analysis suggests a promising role for BSA in improving crop productivity especially for dry land crops, and in acid, poor-structured soils though there was wide variation with soil, crop and biochar properties. Long-term field studies are needed to elucidate the persistence of BSA’s effect and the mechanisms for improving crop production in a wide range of agricultural conditions. At current prices and C-trading schemes, however, BSA would not be cost-effective unless persistent soil improvement and crop response can be demonstrated.