BACKGROUND: Nonunion of long bones is a difficult clinical problem and challenges the clinical acumen of surgeons. Multiple surgical or nonsurgical modalities have been used to treat nonunions. Noninvasive pulsed electromagnetic stimulation is an entity known to affect the piezoelectric phenomenon of bone forming cells. We conducted a study on 45 long-bone fractures of tibia treated by pulsed electromagnetic stimulation, which are analyzed and reported. MATERIALS AND METHODS: A total of 45 tibial fractures with established atrophic nonunion were enrolled between 1981 and 1988. All the patients had abnormal mobility and no or minimal gap at fracture site with no evidence of callus formation across the fracture site. The patients' age ranged between 24 and 68 years; 40 were men and 5 were women. All patients having evidence of infection, implant in situ, and gap nonunions were excluded from study. Pulsed electromagnetic stimulation was given using above-knee plaster of Paris cast (0.008 Weber/m2 magnetic field was created for 12 h/day). The average duration for pulsed electromagnetic stimulation (PEMS) therapy was 8.35 weeks, with the range being 6-12 weeks. The cases were evaluated at 6 weeks and subsequently every 6-weekly interval for clinical and radiological union. The withdrawal of therapy was decided as per clinicoradiological evidence of union. RESULTS: All but three patients showed evidence of union. About 35% (n = 16) cases showed union in 10 weeks, and 85%(n = 38) cases showed union in 4 months. The average duration of therapy using PEMS was 8.35+/-0.48 weeks, and the average duration of immobilization was 3.02 +/- 0.22 months. Three cases that did not show evidence of union were poorly compliant for the apparatus of PEMS. CONCLUSION: PEMS is a useful noninvasive modality of treatment for difficult nonunion of long bones.