The presence of fever and leukocytosis have traditionally been utilized as important diagnostic markers of infection despite some who question their reliability. To examine this point, the role of fever and leukocytosis as diagnostic and prognostic indicators for surgical infections was evaluated. A prospective observational study was performed on all patients with suspected infection in 1997 on the general surgical services at a university hospital. Fever was defined as maximum temperature (Tmax) ≥ 38.5°C, and leukocytosis was defined as a white blood cell (WBC) count ≥ 11, 000/μl. Among all infections, patients presenting with a Tmax ≥ 38.5°C were younger (51.3 ± 1.1 vs. 53.8 ± 0.9 years, p = 0.005) and had a higher APACHE II score (15.1 ± 0.5 vs. 11.4 ± 0.4; p < 0.001). By logistic regression analysis chronic renal insufficiency was associated with a Tmax < 38.5°C [odds ratio (OR) 0.371, 95% confidence interval (CI) 0.195-0.704], and chronic steroid therapy was associated with a WBC count < 11, 000/μl (OR 0.556, 95% CI 0.335-0.921). In addition, infected transplant patients were more likely to present with a Tmax < 38.5°C and a WBC count < 11, 000/μl (OR 0.195, 95% CI 0.075-0.502). Mortality rates for infected patients with a Tmax < 38.5°C or > 38.5°C were 11.6% and 12.9%, respectively (p < 0.7), and the lengths of stay were 14 ± 1 and 18 ± 1 days, respectively (p < 0.03). Mortality rates for patients with a WBC count < 11, 000/μl or > 11, 000/μl were 4.7% and 18.6%, respectively (p < 0.001), and the lengths of stay were 14 ± 1 and 19 ± 1 days, respectively (p < 0.001). In the setting of infection, chronic renal insufficiency and chronic steroid therapy are associated with suppression of fever and leukocytosis, respectively. Transplantation is an independent predictor of infection in patients presenting without fever or leukocytosis. Leukocytosis, but not fever, may be predictive of hospital mortality in infected surgical patients.