Aims:To compare the pharmacodynamics of cisatracurium and rocuronium-induced neuromuscular block following single dose, allowing either spontaneous or neostigmine-accelerated complete recovery.Methods: Following the ethics committee approval and informed consent, 120 patients scheduled for elective surgery under TIVA with tracheal intubation were randomized into 4 groups with different cisatracurium (CIS, 0.10 or 0.15 mg.kg -1 ) or rocuronium (ROC, 0.60 or 0.90 mg.kg -1 ) doses administered. For each patient, the onset time for 95 % depression of T 1 , clinical duration until 25 % recovery, recovery index (T 1 from 25 to 75 %) and time from T 1 25 % to TOF-ratio 0.9 were determined allowing either spontaneous or induced recovery.Results: The onset times were 277 (SD 58), 220 (46), 91 (16) and 77 (16) s for the CIS 0.10, CIS 0.15, ROC 0.60 and ROC 0.90 groups (p < 0.05), respectively, with lower variability in both ROC groups (p < 0.05). The clinical durations were 42 (7), 52 (7), 35 (11) and 52 (12) min, respectively (p < 0.05 for lower doses). Recovery index was identical in all groups allowing either spontaneous recovery -15.9 (1.8), 15.5 (1.7), 16.1 (3.7) and 16.1 (4.0) min, or following neostigmine administration -4.4 (0.9), 4.5 (0.8), 4.3 (0.8) and 4.7 (0.7) min for respective groups. During spontaneous recovery, the variability of DUR25-TOF90 was twice as great for ROC than CIS groups (p < 0.05), while after neostigmine administration it was uniform in all groups.Conclusions: For equipotent doses, the onset times for CIS were approximately three times longer than for ROC. The average clinical duration for both relaxants ranged from 35 to 52 min with acceptable variability. Neostigmine administration accelerated the recovery and reduced its variability. When allowing for spontaneous recovery, less scatter was demonstrated for both CIS groups than for ROC ones.