To investigate if it is appropriate to apply urine density correction when a urine sample is dense or dilute. Data on hippuric acid (HA-U), toluene (Tol-U), creatinine (CR) and specific gravity (SG) in end-of-shift urine samples and exposure to air-borne toluene were cited from previous publications. In practice, 837 cases were available, and they were classified into dense, intermediate and dilute groups taking 0.3 and 3.0 g/l of CR and 1.010 and 1.030 of SG as cut-off points. Lines of regression of HA-U and Tol-U (as observed, CR-corrected or SG-corrected) with air-borne toluene were calculated for each density groups, and correlation coefficients (CCs) were compared. The dense groups gave CCs similar to those of the intermediate groups. Dilute versus intermediate group comparison also gave promising results. These conclusions were however based primarily on the findings with observed values, because the numbers of cases in the dilute or dense group were limited when CR- or SG-correction was applied. Literature survey showed that urine density correction does not always improve the correlation between solvents in air and exposure makers in urine. It was concluded that no correction for urine density may be necessary in evaluating HA-U and Tol-U in dense (and probably also dilute) urine samples as markers of occupational toluene exposure. Just in case when correction for urine density is desired for any reason, SG-correction may be recommended.