The 24-h urine protein-to-creatinine ratio is the gold standard in evaluating proteinuria in lupus nephritis; however, the urine collection is inconvenient to the patient. Random spot urine protein-to-creatinine ratios, although convenient, have poor agreement with the 24-h ratios in these patients. Here, we sought to define a timed collection interval providing accurate and precise data and patient convenience. Urine from 41 patients, in 2 medical centers, with biopsy-proven lupus nephritis was collected at 6-h intervals for 24 h. The protein-to-creatinine ratio of each short collection was then compared with that of a 24-h collection made by combining the 6-h samples. A first morning void and spot urine samples were collected before and after the 24-h collection, respectively. There was significant diurnal variation with peak proteinuria at 6-12 h and nadir at 18-24 h. Each 6-h collection showed excellent correlation and concordance with the 24-h protein-to-creatinine ratio, but the 12-24-h interval had the best agreement. In contrast to the random spot urines, the first morning void also had excellent correlation and concordance, but underestimated the 24-h protein-to-creatinine ratio. Our study shows that a 12-h overnight urine collection is the best surrogate, with excellent agreement with the 24-h protein-to-creatinine ratio, and it is convenient for patients. There was little variability between centers, an important feature for clinical trials.