A study where researchers observe outcomes without intervening or assigning treatments. Provides weaker evidence than RCTs due to potential confounding factors, but useful for studying long-term safety.