OBJECTIVE To determine the impact of administration rate and buffering on the pain associated with subcutaneous infiltration of lidocaine. METHODS Forty-two adult volunteers employed at a tertiary care center participated in this prospective, single-blinded study. Each subject received four lidocaine injections prepared and administered as follows: slow, buffered (SB); slow, unbuffered (SU); rapid, buffered (RB); rapid, unbuffered (RU). Buffering was accomplished by mixing 1% lidocaine with 8.4% sodium bicarbonate in a 9:1 ratio. Slow administration was 30 seconds and rapid was 5 seconds. Needle size (27-gauge), injection depth (.25 inch), lidocaine volume (1.0 mL), and temperature (room) were the same for each of the four injections. In all four conditions, the needle remained in the forearm for 30 seconds, to ensure blinding. The main outcome measure was the mean pain score for each condition, as recorded on a 10-cm visual analog scale. RESULTS The lowest pain scores (mean +/- SE) were recorded for the SU and SB conditions at 1.49 +/- 29 and 1.48 +/- 26, respectively, and they were significantly lower than the scores for RB (2.34 +/- 28; P < .01) or RU (3.11 +/- 33; P < .001). Each of the slow conditions was reported to be the "least painful" of the four significantly more often than either rapid condition. CONCLUSIONS This is the largest blinded study to assess administration rate and the pain of a local anesthetic. We found that administration rate had a greater impact on the perceived pain of lidocaine infiltration than did buffering.