Background Regular physical activity (including structured exercise) reduces the risk of developing site-specific cancers, including breast, endometrial, and colon cancer. However, the underlying biological mechanisms are not fully understood. During exercise, skeletal muscle and other secretory organs release cytokines and other peptides into the circulation, which have the potential to influence key signalling pathways involved in cancer progression. Aims To determine the effect of exercise-conditioned serum on cancer cell growth in vitro, we: i) systematically reviewed and meta-analysed the available evidence, and ii) stimulated a colon cancer cell line with serum collected immediately pre- and post-exercise from adults with lifestyle risk factors for colon cancer.   Methods Five literature databases were systematically searched for studies that assessed cancer cell growth after exposure to human serum collected before and immediately after an acute bout of exercise. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were pooled using a three‐level random‐effects model. Meta‐regressions were performed with participant age and disease status, exercise type, cell line phenotype, and serum incubation time as covariates. In addition, we recruited physically inactive, overweight (BMI ≥25 kg/m2) males aged ≥50 years into a randomised, controlled, crossover study. Participants completed an acute bout of moderate-intensity aerobic interval exercise (6 x 5-min @ 60% heart rate reserve) and a non-exercise control condition (60-min rest) in a randomised, counterbalanced order. Serum samples were collected immediately before and after both conditions and used to stimulate a human colon cancer cell line (LoVo). Colon cancer cell growth was evaluated after 48 hours of incubation by the resazurin assay. The difference in cell growth was assessed using a multilevel linear model, with change from baseline as the dependent variable, baseline values as a covariate, condition (exercise or control) as a fixed effect, and participants as a random factor. Results Seven studies encompassing 98 participants were included in the meta-analysis. Studies used a range of cancer cell lines including breast, colon, lung, and prostate. Pooled analysis showed that exercise‐conditioned serum reduced cancer cell growth compared with pre-exercise serum (SMD = −1.23, 95% CI: −1.96 to −0.50; I2 = 75.1%). However, the certainty of evidence was very low due to unexplained heterogeneity, high probability of publication bias, and risk of bias within individual studies. Sixteen participants were recruited into the crossover study (age: 60.0 ± 8.0 years; BMI: 29.9 ± 2.4 kg/m2). The median heart rate reserve during aerobic interval bouts was 59.7% (IQR: 57.6 to 61.8%). Incubating colon cancer cells with exercise-conditioned serum reduced cancer cell growth compared with the non-exercise control condition (MD: -5.7%, 95% CI: -8.8 to -2.6%) (Figure 1).  Conclusions The results of our meta-analysis and crossover study provide evidence that exercise‐conditioned serum reduces cancer cell growth in vitro. This suggests that acute exercise‐induced modulations in serum could contribute to the reported inverse associations between physical activity and cancer risk. However, many questions remain regarding the underlying mechanistic pathways, relevance of in vitro findings to in vivo physiology, and potential effect modifiers such as exercise intensity and cancer cell phenotype.