OBJECTIVE Periprostatic nerve block (PPNB) is the standard anesthesia for ultrasound (US) guided transrectal prostate biopsy (TPB), but periprostatic infiltration itself constitutes a major, though often neglected, source of discomfort even in patients receiving perianal-intrarectal lidocaine-prilocaine (PILP) cream before PPNB. Noninfiltrative anesthesia therefore represents an attractive alternative to periprostatic infiltration. With this in mind, we aimed to determine the efficacy and safety of perianal-intrarectal (PI) lidocaine gel, lidocaine-ketorolac gel, and lidocaine-prilocaine cream in relieving pain during TPB. METHODS Three hundred consecutive patients scheduled for US-guided TPB were randomized 1:1:1 to receive PI administration of 5 g 2.5% lidocaine gel 10 minutes before TPB (Group 1), or a mixture of 5 g 2.5% lidocaine gel and 0.3% ketorolac tromethamine solution 1 hour before TPB (Group 2), or 5 g 2.5% lidocaine and 2.5% prilocaine cream 20 minutes before TPB (Group 3). The 0-to-10 points visual analogue scale (VAS) was used for assessing pain at probe insertion and movements (VAS-1), at prostate sampling (VAS-2), and maximal procedural pain (MPP). Complications occurring up to 20 days after the procedure were also recorded. RESULTS Four (1.3%) patients were excluded because of unbearable pain during the procedure, leaving Group 1 with 98 patients, Group 2 with 99, and Group 3 with 99; the 3 groups were comparable for patients' age, serum PSA, prostate volume, and cancer detection rate. The addition of either ketorolac or prilocaine to lidocaine significantly (P < 0.0001) reduced probe-related, sampling-related, and maximal procedural pain. Compared with lidocaine-prilocaine, lidocaine-ketorolac was less effective in relieving probe-related pain (mean VAS-1: 1.47 ± 1.30 vs. 0.39 ± 0.65; P < 0.0001) but more effective in relieving sampling-related pain (mean VAS-2: 0.76 ± 0.94 vs. 1.54 ± 1.02; P < 0.0001); there was no difference in MPP (mean 1.82 ± 1.21 vs. 1.67 ± 0.95), probably due to such different efficacy on the two pain sources. Complications were similar in the 3 groups. CONCLUSIONS Lidocaine-prilocaine cream was most effective on probe-related pain, whereas lidocaine-ketorolac gel was most effective on sampling-related pain. These noninfiltrative anesthetics were safe, easy to administer, and well accepted by patients; the possibility to combine them to further improve pain control during TPB deserves further well-designed studies.