Previously, prior approval from the on-call chemical pathologist was required in our hospital for plasma paracetamol measurements. However, since May 1992, there have been no restrictions on ordering this assay. We have assessed the consequences of this policy change by comparing the number and appropriateness of requests for plasma paracetamol measurements in Chinese patients admitted to our hospital with acute poisoning over two six-month periods (July-December) in 1991 and 1993. Requests were considered appropriate if paracetamol ingestion was suspected or unknown drugs were ingested. The number of patients having plasma paracetamol concentrations assayed increased from 51 in 1991 to 141 in 1993 (176%). The corresponding increase in the number of Chinese patients admitted to two of our eight general wards with poisoning was estimated to be 93%. The proportion of 'appropriate' plasma paracetamol measurements dropped from 55% in 1991 to 21% in 1993. Eight patients had plasma paracetamol concentrations above the recommended treatment line; they were all from the group in whom the requests were appropriate. Three of the 135 patients in the group with 'inappropriate' requests were found to have slightly elevated but far from toxic plasma paracetamol concentrations. Unrestricted availability of plasma paracetamol measurements resulted in an increase in the number of inappropriate requests.