Bone and gallium scans are used to assess osteomyelitis patients with prior bone disease. To refine the criteria for interpreting these scans, the data from 136 consecutive patients with clinically suspected osteomyelitis were reviewed. Active osteomyelitis was diagnosed with surgery or biopsy and culture in 49 patients, excluded with the same criteria in 16, and excluded by clinical follow-up for at least 6 months in 71. Five different scintigraphic patterns were found. The true-positive and false-positive ratios, the likelihood ratios, and posterior probabilities for active osteomyelitis in each pattern were calculated. Only one pattern (gallium uptake exceeding bone-seeking radiopharmaceutical uptake) was indicative of active disease. Other patterns slightly raised or decreased the probability of disease. The extent of these changes varies directly with the prior probability of disease, determined from patient-specific factors (e.g., clinical data, laboratory data, findings on plain films) known best by the referring clinician.