The active morphine metabolite, morphine-6-glucuronide (M-6-G), may contribute to both the analgesia and the adverse effects observed during morphine (MOR) therapy. To evaluate the relationship between M-6-G and adverse effects, we measured steady-state plasma concentrations of MOR and M-6-G and concurrently noted the presence or absence of moderate to severe cognitive impairment or myoclonus in 109 cancer patients who were receiving either oral (n = 71) or parenteral (n = 38) morphine. MOR and M-6-G plasma concentrations were determined by HPLC with electrochemical detection. The presence of cognitive impairment or myoclonus was analyzed in relation to molar M-6-G/MOR ratio, age, morphine dose, the use of other centrally acting drugs, renal function (blood urea nitrogen (BUN) and serum creatinine), hepatic function (serum bilirubin, serum glutamic oxalacetic transaminase (SGOT), and alkaline phosphotase) and serum lactate dehydrogenase (LDH). The patient population consisted of 60 women and 49 men. The mean age was 51.5 years (range: 10-85 years). The mean morphine dose (total dose-prior 48 h) was 486 mg (range: 40-4800 mg) for the oral group and 931 mg (range: (10-9062 mg) for the parenteral group. The mean molar M-6-G/MOR ratios were 6.1 (SD: 18.2; range: 0.01-153.3) for the oral treatment group and 2.7 (SD: 4.16; range: 0.05-23.8) for the parenteral treatment group. Overall, the M-6-G/MOR ratio demonstrated a moderate but significant correlation with BUN (r = 0.4; P < 0.001) and creatinine (r = 0.45; P < 0.001) levels, but not with the other clinical variables examined. Although the prevalence of myoclonus among patients receiving oral morphine was 3-fold higher than those receiving parenteral morphine (P < 0.05), multivariate analyses demonstrated that neither myoclonus nor cognitive impairment was significantly associated with M-6-G/MOR ratio when adjusted for other variables. In contrast, there were significant associations between increasing age, elevated bilirubin or LDH levels, and the presence of cognitive impairment. Finally, analysis of a small subgroup of patients with very high M-6-G concentrations (> 2000 ng/ml) and presenting with either respiratory depression or obtundation, suggested that elevated M-6-G levels were associated with these severe adverse effects primarily in the setting of metabolic dysfunction. Together, these data confirm the previously reported correlation between deteriorating renal function and increasing M-6-G/MOR ratio, but do not support the conclusion that increasing ratio alone is a determinant of two prevalent opioid-related adverse effects: myoclonus and cognitive impairment.