OBJECTIVE: To compare short-term complications of amniocentesis using 20G versus 22G needle.METHODS: A total of 200 women referred for mid-trimester amniocentesis were randomized to a 20G (Group I, n = 100) or 22G amniocentesis needle (Group II, n = 100). The primary outcome was intrauterine bleeding at needle insertion. The operator reported technical aspects and patient's reactions immediately after the procedure. Women's perception of discomfort 30 min after the procedure and complications after 2 weeks were recorded.RESULTS: Intrauterine bleeding at needle insertion was similar between groups (4/100 vs 8/100). When only transplacental taps were analyzed, bleeding was significantly lower in Group I (4/20 vs 8/14, p = 0.035). Fluid retrieval was faster in Group I (9.6 vs 26.8 sec, p < 0.001). In all, 65% of women in Group I versus 30% in Group II reported discomfort during the procedure, although discomfort 30 min after the procedure and complications within 2 weeks after the procedure were similar in the two groups.CONCLUSION: Amniocentesis with 20G needle is associated with lower risk of intrauterine bleeding in case of transplacental needle insertion and allows for faster fluid retrieval, as compared with 22G needle. Nevertheless, 20G needle is associated with more immediate discomfort during the procedure.