As the results of clinical trials are inconsistent, we conducted this research to assess the effect of purified anthocyanins or anthocyanin-rich extract supplementation on C-reactive protein (CRP) levels. We searched several databases to identify and extract data on characteristics, methods and outcomes of the eligible randomised controlled trials (RCT). A random-effects model, weighted mean difference (WMD) and 95 % CI were applied for data analysis. To investigate the effects of the study quality score or design on our results, we performed the same analysis by excluding the studies of Karlsen et al., with the lowest quality score, and Hassellund et al., with a cross-over design. Meta-analysis showed that anthocyanins had no significant impact on CRP levels (WMD=0·018; 95 % CI –0·44, 0·47; P=0·94). Although the effect of anthocyanins was independent of supplementation duration (slope: 0·01; 95 % CI –0·002, 0·03; P=0·08), their effect depended on the dose of anthocyanins (slope: 0·001; 95 % CI 0·0007, 0·002; P<0·001). However, no significant relationship was found between the anthocyanin dosage and CRP levels after excluding the studies of Karlsen et al. and Hassellund et al. Finally, anthocyanins had no effect on CRP levels regarding healthy participants, patients and types of anthocyanins. Although changes in CRP concentrations had no association with trial duration, a significant relationship was found between anthocyanin dosage and CRP level. No significant result was observed between the anthocyanin dosage and CRP levels after excluding the mentioned studies. Further well-designed RCT are needed to validate these findings.