10,000 seems to be a good guideline, e.g., The p-values of bootstrap samples that are larger than 10,000 will be within 0.01 percent of the actual p-value for the method approximately 95% of times.
Disclaimer.
All the information on this website is published in good faith and for general information purpose only. Watson Media does not make any warranties about the completeness, reliability, and accuracy of this information.