This study recruited 20 patients who had undergone modified tension band wiring for patellar fracture, with a mean follow up of 30 months. Subjects were grouped according to results of Cybex isokinetic testing at 60 degrees /s angular velocity. Subjects with <30% deficit as compared with the contralateral knee constituted group I, and those with >30% similar deficit formed group II. Plain radiography and HSS scoring were also performed. According to patient satisfaction, HSS scoring and Cybex testing, results were good in 80%, 90% and 55% of cases, respectively. Statistical analysis revealed that there were no significant differences between the two groups in terms of age, gender, duration of follow up, fracture type (two-part or comminuted), dominancy, time interval between trauma and surgery, or duration of immobilisation. The number of patients with >1 mm articular incongruity postoperatively was significantly higher in group II, which also had significantly higher incidences of >1 cm thigh atrophy and pain and thus increased deficits. We suggest that articular incongruity should be limited strictly to 1mm in surgery for patellar fractures.