Certain roentgenographic signs have value in predicting the fixation of a cementless femoral component to bone by osseointegration. Other signs have value in predicting the gross stability of a cementless femoral component. The authors have determined the specificity and sensitivity of the signs for osseointegration in cases in which the histologic fixation has been confirmed after implant removal. The authors have also determined the specificity and sensitivity of the signs for gross implant stability in cases in which the stability has been confirmed at reoperation. Statistical methods were used to determine a numeric value for each of these roentgenographic signs, and these values were combined into a score. The score was divided into fixation by osseointegration and mechanical implant stability. The two scores were then combined into an overall score. When signs of osseointegration were present, the implant was always stable, and the overall scores were the highest. When signs of osseointegration were absent, the mechanical stability varied, and the scores were lower. A neutral or slightly negative score correlated with failed osseointegration but secondary successful implant stabilization. A very low negative score correlated with gross implant instability. To confirm the validity of the scoring system, the two-year postoperative score was determined for 1005 cases in which the clinical outcome was known. A strong correlation between the presence of symptoms and a low score confirmed the value of the scoring system for diagnosing implant loosening as the cause of symptoms. As a second test, the two-year and five-year postoperative results were compared in the same patients. A high two-year score correlated with durable implant stability through five years. A low two-year score correlated with a higher incidence of late symptomatic loosening.