Objective To explore the value of procalcitonin (PCT) in differential diagnosis of invasive candidiasis. Methods PubMed, Embase, Medline, Cochrane Library, China National Knowledge Infrastructure and Wanfang Data were searched for articles published from the dates of establishment of databases to January 2021. A prospective and retrospective cohort studies and a case-control studies of PCT in differential diagnosis of invasive candidiasis were collected. RevMan 5.3 software QUADAS-2 risk assessment tool was used to evaluate the quality of the literature. Meta-Disc 1.4 software was used to determine whether the original data had threshold effect and heterogeneity. Stata 14.0 software was used to analyze meta, judge publication bias and draw Deeks diagram. Results A total of 9 articles and 943 patients were included. There were 259 cases in candida group and 684 cases in control group. The study showed that the total sensitivity was 0.86 [95% confidence interval (CI) (0.80, 0.91)], specificity was 0.78 [95%CI (0.70, 0.84)], positive likelihood ratio was 3.92 [95%CI (2.77, 5.55)], negative likelihood ratio was 0.18 [95%CI (0.12, 0.27)], the area under receiver operator characteristic curve was 0.90 [95%CI (0.87, 0.92)], diagnostic odds ratio was 19.75 [95%CI (10.71, 36.43)]. The results of heterogeneity test showed that there was heterogeneity caused by non-threshold effect between studies. The results of subgroup analysis showed that the heterogeneity I2 value of PCT<2 ng/mL subgroup decreased significantly, and the result was more stable, with sensitivity. The results show that sensitivity was 0.86 [95%CI (0.79, 0.91)], specificity was 0.72 [95%CI (0.63, 0.80)], the area under receiver operator characteristic curve was 0.87 [95%CI (0.83, 0.89)]. Conclusions Serum PCT in the differential diagnosis of invasive candidiasis has certain accuracy and negative predictive value. However, PCT is only an auxiliary test. The differential diagnosis of invasive candidiasis should be combined with clinical features and other diagnostic indexes.