Given that pain relief is often the primary goal of orthopaedic surgery, an accurate assessment of pain is paramount. The objectives of this cross-sectional analytical study were to (1) compare how the Patient-Reported Outcomes Measurement Information System (PROMIS) pain interference (PI) computer adaptive test (CT) performs against the Numeric Pain Scale (NPS) measure in evaluating pain, and (2) to determine demographic, clinical, and psychosocial correlates of PI in an urban population undergoing a variety of knee surgeries. We hypothesized that there would be a strong correlation between PI and NPS, with minimal floor and ceiling effects; and that a worse PI score would be associated with a worse general health profile. The sample consisted of 412 patients undergoing knee surgery at an urban academic center. Patients were preoperatively administered measures of health-related quality of life (HRQOL). Bivariate and multivariable statistical analyses were performed to identify significant independent predictors. The mean PI score was 60.3 ± 7.2 and had no floor or ceiling effects, whereas NPS demonstrated a greater percentage of patients scoring at the extremes of the measure. Worse PI scores were associated with older age, higher body mass index (BMI), greater comorbidity, lower income, smoking, female gender, Hispanic ethnicity, Black race, unemployment, opioid use, lower expectations, and greater American Society of Anesthesiologists score (p < 0.05). Compared with other procedures, total knee arthroplasty was associated with worse PI scores and anterior cruciate ligament reconstruction was associated with better PI scores. Furthermore, PI demonstrated significant associations with a wide range of HRQOL measures. After controlling for confounding variables, worse PI was independently associated with older age, lower income, higher BMI, and smoking.