The aim of this study was to provide a comprehensive comparison of training modalities for countermovement jump (CMJ) performance in soccer players using a network meta-analysis (NMA). PubMed, Web of Science, Embase, EBSCO, and the Cochrane Library were searched up to November 2025. Randomized controlled trials (RCTs) involving healthy soccer players, evaluating structured training interventions against comparator training or control conditions, and reporting CMJ height outcomes were included. A random-effects NMA was performed to estimate mean differences (MDs) with 95% confidence intervals (CIs), and interventions were ranked using P-scores, integrating both direct and indirect evidence. Subgroup analyses were conducted according to age, sex, and competitive level. Twenty-nine RCTs (n = 450) were included. The network showed moderate heterogeneity and no evidence of significant inconsistency. Combined training (CBT) was ranked highest in effect estimates in CMJ performance compared with traditional soccer training (TST) (MD = 4.92, 95% CI: 3.02-6.83, P-score = 0.98), followed by flywheel resistance training (FT) and plyometric training (PT), whereas other interventions did not demonstrate significant improvements. Subgroup analyses indicated that CBT ranked highest in males, youth, and Tier 2 players; PT ranked highest in females and adults; and FT ranked highest in highly trained players. Given the limited sample sizes, limited statistical power, and the fragmented nature of some subgroup comparisons, these findings should be interpreted cautiously. The network showed moderate heterogeneity (I² = 32.4%) and no evidence of significant inconsistency. The certainty of evidence was generally low to moderate according to the CINeMA assessment. Directness of evidence was generally high across comparisons, although several key contrasts (e.g., CBT vs FT and FT vs PT) relied partly on indirect evidence within the network. Although the highest-ranked interventions (CBT, FT, and PT) were supported primarily by direct comparisons with TST, several active–active contrasts relied partly on indirect estimates. The reliability of the intervention rankings depends on similarity in participant characteristics, intervention dose, outcome measurement, and control conditions.

• This network meta-analysis synthesized 29 RCTs involving 450 soccer players and compared eight training modalities for improving countermovement jump performance.

• Combined training ranked highest overall compared with traditional soccer training, followed by flywheel resistance training and plyometric training.

• Exploratory subgroup analyses indicated that combined training ranked highest in males, youth, and Tier 2 players; plyometric training in females and adults; and flywheel resistance training in Tier 3 players.

• The certainty of evidence was generally low to moderate, and the limited number of direct head-to-head comparisons between active interventions warrants cautious interpretation of treatment rankings.