Request throttling is the deliberate practice of slowing down the rate at which a scraper sends requests, adding delays between them, to mimic natural human browsing patterns and stay under a target site's rate limits. Throttling trades speed for reliability, reducing the odds of triggering rate limits or bot-detection systems.

Throttling can use randomized rather than fixed delays between requests (Scrapy's DOWNLOAD_DELAY_JITTER setting, for example, adds random variation to each delay). Throttling and proxy rotation work together rather than as substitutes: rotation spreads load across many IPs, while throttling controls how aggressively each individual IP is used.

Throttling can also adapt to a site's responses, as Scrapy's AutoThrottle extension does by adjusting delays to server latency and not speeding up on error responses, rather than relying on a fixed delay chosen in advance.