OBJECTIVES: To extend well-established methods for sample size in power calculations in pair wise meta-analysis to the scenario where multiple treatments are being analyzed in a network meta-analysis METHODS: We derive methods of approximating the ‘effective number of patients’ in indirect comparison meta-analysis where no head-to-head evidence is available. We calibrate these approaches with conventional approaches for estimating the required sample size and power in pair wise meta-analysis.  RESULTS: The calibration of the above two methods allow for a simple assessment of the power and strength of evidence for each treatment comparison in a network of treatments. The resulting measures are 1) the statistical power associated with each treatment comparison made in a network meta-analysis; 2) the effective number of patients for each comparison contrasted, which can be contrasted with the required sample size for the particular comparison to gauge the strength of evidence. We provide an illustrative example using data from a network meta-analysis of interventions for smoking cessation. CONCLUSIONS: The proposed measured follow the format of well-known sample size and power measures. They are easy to calculate and will resonate with a statistically non-sophisticated audience.