Compliance with 24-h urine samples can be low in population-based studies and first morning void urine samples are often collected for convenience. Interpretation of arsenic concentrations in urine is influenced by a range of factors unrelated to exposure. To reduce the influence of such factors, creatinine adjustment is routinely used. This study aimed to determine whether first morning void urinary inorganic arsenic concentrations approximate 24-h urinary arsenic concentrations and whether creatinine adjustment improved the correlation between environmental arsenic concentrations and urinary inorganic arsenic concentrations. One hundred sixty spot samples and corresponding 24-h urine samples were collected from people living in areas with a range of environmental arsenic concentrations and analyzed for inorganic arsenic using borohydride arsine generation followed by atomic absorption spectrophotometry. There were no significant differences between the urinary inorganic arsenic concentrations for the different sample types or whether creatinine adjusted or not. Significant correlations were observed between adjusted and unadjusted samples. The data set was highly skewed and when only detectable arsenic samples were considered, the relationship between sample types became nonsignificant. The results of this study suggest that spot samples may be adequate for measuring short-term exposure, using inorganic arsenic as the outcome variable; however, additional work on a larger data set is required. Creatinine adjustment of urinary inorganic arsenic concentrations may not be required in population studies investigating environmental exposure. (C) 2002 Elsevier Science (USA).