The qualitative direct barbituric acid (DBA) method of detecting urine nicotine metabolites was modified to make it quantitative. The performance of the quantitative DBA method was compared with the qualitative method and an established cotinine radio-immunoassay (RIA), using a panel of urines from 128 reported smokers and 383 reported non-smokers. The quantitative DBA method results were highly correlated with the cotinine RIA results, r = 0.85. The coefficients of variation for the two methods were 6% and 10%, respectively. Assuming that the reported smoking history was correct the qualitative DBA method gave a smoking detection rate of 91% and a false positive rate of 3%. At cut-off levels chosen to yield the same false positive rate the quantitative DBA method detected 93% of smokers, close to that of 98% detected with the cotinine RIA. The quantitative DBA method can be used to analyse over 170 samples per day compared to about 70 per day by RIA. It is therefore a fast and inexpensive alternative to cotinine assays for the assessment of smoking status and cigarette smoke intake.