CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / Os valores de referÃncia de exames laboratoriais sÃo definidos com base em critÃrios estatÃsticos para uma amostra aleatÃria de indivÃduos. Os parÃmetros de referÃncia publicados sÃo provenientes de uma variedade de amostras, incluindo doadores de sangue, participantes de exames admissionais e voluntÃrios normais que fazem exames de rotina para ensaios clÃnicos. Esta pesquisa à do tipo documental com abordagem quantitativa. A amostra foi constituÃda por 1.947 exames laboratoriais dos voluntÃrios sadios, de ambos os sexos, com idade mÃdia de 25 anos, que participaram de ensaios clÃnicos na UNIFAC, entre os anos de 1999 e 2003. Para a realizaÃÃo do estudo, foram colhidos dos prontuÃrios as seguintes informaÃÃes: dados de identificaÃÃo do voluntÃrio, alÃm dos parÃmetros hematolÃgicos e bioquÃmicos. Foram investigados setenta e cinco ensaios clÃnicos, tendo os resultados demonstrado diferenÃa estatÃstica entre os sexo, para a maioria dos parÃmetros analisados. As faixas de referÃncia foram calculadas pelo mÃtodo dos percentis, sendo encontrado os seguintes valores para a anÃlise hematolÃgica para mulheres e homens, respectivamente: eritrÃcitos - de 3,9 a 5,0 milhÃes/mm3 e 4,4 a 5,7 milhÃes/mm3; hemoglobina - de 14,8 a 15,3 g/dL e 13,6 a 16,9 g/dL; leucÃcitos - de 4.400 a 10.500/ mm3 e 4.200 a 10.100/mm3; plaquetas â de 165.000 a 397.000/ mm3 e 155600 a 354400/ mm3. Jà para os parÃmetros bioquÃmicos, verificaram-se os seguintes valores para mulheres e homens, respectivamente: creatinina - de 0,5 a 0,9 mg/dL e 0,7 a 1,2 mg/dL; glicose - de 69 a 96 mg/dL e 71 a 100 mg/dL; colesterol - de 118 a 224 mg/dL e 112 a 219 mg/dL; triglicerÃdeos - de 38,0 a 171 mg/dL e 42 a 197 mg/dL. Os nossos achados sugerem que os valores de referÃncia dos parÃmetros investigados, para a populaÃÃo metropolitana de Fortaleza, sejam redefinidos como aqueles incluÃdos no intervalo entre os percentis 2,5o e 97o, os quais correspondem a 95% da distribuiÃÃo total dos valores de cada parÃmetro. / The values of reference of laboratorial parameters are defined on the basis of statistical criteria for a random sample of individuals. The published parameters of reference are proceeding from a variety of samples, including blood donors, participants of job admission examinations and normal voluntaries that are submitted to routine examinations for clinical assays. This research is of the documentary type with quantitative approach. The sample was constituted by 1.947 laboratorial parameters of healthy volunteers, of both genders, with average of 25 years, which had participated of clinical assays in the Clinical Pharmacological Unit (UNIFAC â UFC), between the years of 1999 and 2003. For the accomplishment of the study, the following information had been obtained from case report files: identification of the volunteer, besides the hematological and biochemical parameters. Seventy five clinical assays have been investigated. The results demonstrated statistical differences between the genders, for the majority of the analyzed parameters. The normal reference intervals have been calculated by the method of percentiles, being found the following values for the hematological analysis for women and men, respectively: erytrocytes â from 3.9 to 5.0 millions/mm3 and from 4.4 to 5.7 millions/mm3; hemoglobin â from 14.8 to 15.3 g/dL, and from 13.6 to 16.9 g/dL; leukocytes â between 4.400 and 10.500/mm3, and between 4.200 and 10.100//mm3; platelets â from 165.000 to 397.000/mm3, and from 155.600 to 354.400/mm3. For the biochemical parameters, there were found the following values for women and men, respectively: creatinine â from 0.5 to 0.9 mg/dL, and from 0.7 to 1.2 mg/dL; glucose â between 69 and 96 mg/dL, and between 71 and 100 mg/dL; cholesterol â between 118 and 224 mg/dL, and between 112 and 219 mg/dL; triglycerides â from 38 to 171 mg/dL, and from 42 to 197 mg/dL. Our findings suggest that the reference values of investigated parameters, for the metropolitan population of Fortaleza, must be redefined as that enclosed in the interval between the 2.5 and 97, which correspond to 95% of the total distribution of the values of each parameter.