Causal links between vitamin D status [25(OH)D] and systemic inflammation were examined through a systematic review of randomized controlled trials (RCTs). Selected RCTs were ⩾12 weeks, conducted in adults free of acute inflammatory disease, and of high-quality (Jadad score ⩾3). Of 14 studies that met our criteria, 9 studies (15 study arms) permitted extraction of data. There was no effect on the weighted mean difference (WMD) of IL-6 (WMD (95% confidence interval)=0.1, (-0.166, 0.366) pg/ml, P=0.462) or C-reactive protein (CRP) (WMD=-0.324, (-1.007, 0.359) mg/l, P=0.352). Subgroup analyses of trials achieving ⩾80 nmol/l indicated a trend for lower CRP (WMD=-0.834, (-1.726, 0.058) mg/l, P=0.067), however heterogeneity was significant (I2=66.7%, P=0.017). Studies employing a low dose (<1000 IU/d) showed increased CRP (WMD=0.615, (0.132, 1.098), P=0.013). In contrast, ⩾1000 IU/d had a favourable effect on CRP (WMD=-0.939, (-1.805, -0.073), P=0.034) but heterogeneity was significant (I2=61.3%, P=0.017). Meta-regression indicated that older age predicted a significant decrease in IL-6 (β=-0.02, (-0.034, -0.006) pg/ml, P=0.013) and CRP (β=-0.06, (-0.103, -0.017), P=0.01), whereas a greater percentage of females (β=0.027, (0.011, 0.044), P=0.004) and longer study duration independently predicted a higher WMD for CRP (β=0.049, (0.018, 0.079), P=0.005). Available high-quality RCTs did not support a beneficial effect of cholecalciferol on systemic IL-6 and CRP. Future studies should consider the confounding effects of age, gender and study duration, while possibly targeting an achieved 25(OH)D ⩾80 nmol/l.