We examined the usefulness of thiosulfate as an indicator of hydrogen sulfide poisoning by analysing sulfide and thiosulfate in three cases. In the first (non-fatal) case sulfide and thiosulfate were not detected in the blood samples from any of the four workers involved in the accident. In the urine samples, only thiosulfate was detected in three out of the four workers at a concentration of 0.12-0.43 micromol/ml, which was 4-14 times higher than the level in a healthy person. In the second (fatal) case sulfide and thiosulfate were detected in the blood sample at concentrations of 0.007 micromol/ml for sulfide, and 0.025 micromol/ml for thiosulfate. The thiosulfate concentration was at least 8 times higher than the level in a healthy person. In the third (fatal) case sulfide and thiosulfate were detected in the blood sample at concentrations of 0.95 micromol/ml for sulfide, and 0.12 micromol/ml for thiosulfate. Based on the above results, we concluded that thiosulfate in urine is the only indicator to prove hydrogen sulfide poisoning in non-fatal cases, while the analysis of sulfide in fatal cases should be accompanied by the measurement of thiosulfate in blood.