This study investigates the influence of patellar thickness in total knee arthroplasty (TKA) with routine patellar resurfacing on the rate of lateral retinacular release. A study group comprised of 121 TKAs using surgical instrumentation allowing a measured resection of the patella was compared with a control group comprised of 100 TKAs using an "eye-ball" technique for patellar resurfacing without particular attention to patellar thickness. A composite without patellar thickness equal to or slightly less than the original patella was attempted in the study group. The average thickness difference of the patella in the study group was -1.58 mm. Lateral retinacular release for patellar maltracking was performed in 12.4% of knees in the study group compared with 55% of knees in the control group. The variation in patellar thickness in this series often led to a residual bone thickness < 15 mm. No clinical problems have been observed to date.