To evaluate the 24-hour urine protein-to-creatinine (U pr:cr) ratio compared to 24-h urine total protein excretion as a measure of proteinuria in patients with lupus nephritis. Proteinuria in 8 patients with lupus nephritis treated with cyclophosphamide was monitored by total protein excretion and U pr:cr ratio in 24-h urine collections. A median of 16 measurements per patient were collected over a median of 47 months. Adequacy of the 24-h collection was assessed by comparing total urine creatinine to the predicted creatinine. Collections in which the difference between the predicted 24-h urine creatinine and the measured 24-h urine creatinine was greater than or equal to 20% were defined as inadequate collections. Timed 24-h urine collections were frequently inadequate (30.2% of total collections were under-collections, while 14.3% were over-collections). We found 87.5% of patients had at least one under-collection whereas 75% had at least one over-collection. Correlations between total protein and U pr:cr ratio for individual patients ranged from 0.87 to 0.99 (mean 0.95). For the entire sample, the correlation (R2 = 0.89) of total urine protein to Upr:cr ratio was excellent. Excluding the 38 under-collections led to improvement in the overall correlation (0.94). Excluding the 18 over-collections led to a correlation of 0.89. Excluding both under-collections and over-collections led to a correlation of 0.94. In patients with lupus nephritis, the 24-h U pr:cr ratio is highly correlated with the 24-h urine protein excretion when the collections are adequate. The error of the estimate is higher when collections are poor.