It is difficult to assess outcomes in randomized controlled trials in SLE. Its cyclic nature and multiorgan involvement complicate the identification and reproducibility of outcome measures. Disease activity indices do not necessarily reflect disease outcome in the generally short timeframes of protocol treatment. Patient and physician assessments of disease may not correlate well. Responder analyses do not function well if they are proposed in the absence of data from RCTs. Changes in medical practice may affect patient selection and make it difficult to replicate findings in a subsequent protocol. Background therapy even if stable during protocol participation may confound treatment effects. While objective outcome measures are well defined in renal and hematologic disease, few SLE patients have isolated organ system involvement at any one time. Protocol designs must anticipate changes in disease activity and need for treatment of emergent disease manifestations. Despite these difficulties, a body of evidence derived from RCTs has developed. Although limited and yet to result in an approved therapy, early markers of treatment response have been defined and shown to correlate with longer term clinical outcomes. Biomarkers are an appealing idea to assess biologic effects of study treatment and hopefully predict clinical outcome.