Background: Effects of ketone supplements as well as relevant dose-response relationships and time effects on blood β-hydroxybutyrate (BHB), glucose and insulin are controversial. Objective: This study aimed to summarize the existing evidence and synthesize the results, and demonstrate underlying dose-response relationships as well as sustained time effects. Methods: Medline, Web of Science, Embase, and Cochrane Central Register of Controlled Trials were searched for relevant randomized crossover/parallel studies published until 25th November 2022. Three-level meta-analysis compared the acute effects of exogenous ketone supplementation and placebo in regulating blood parameters, with Hedge's g used as measure of effect size. Effects of potential moderators were explored through multilevel regression models. Dose-response and time-effect models were established via fractional polynomial regression. Results: The meta-analysis with 327 data points from 30 studies (408 participants) indicated that exogenous ketones led to a significant increase in blood BHB (Hedge's g = 1.4994, 95% CI [1.2648, 1.7340]), reduction in glucose (Hedge's g = −0.3796, 95% CI [-0.4550, −0.3041]), and elevation in insulin of non-athlete healthy population (Hedge's g = 0.1214, 95%CI [0.0582, 0.3011]), as well as insignificant change in insulin of obesity and prediabetes. Nonlinear dose-response relationship between ketone dosage and blood parameter change was observed in some time intervals for BHB (30–60 min; >120 min) and insulin (30–60 min; 90–120 min), with linear relationship observed for glucose (>120 min). Nonlinear associations between time and blood parameter change were found in BHB (>550 mg/kg) and glucose (450–550 mg/kg), with linear relationship observed in BHB (≤250 mg/kg) and insulin (350–550 mg/kg). Conclusion: Dose-response relationships and sustained time effects were observed in BHB, glucose and insulin following ketone supplementation. Glucose-lowering effect without increasing insulin load among population of obesity and prediabetes was of remarkable clinical implication. Registry and registry number: PROSPERO (CRD42022360620).