OBJECTIVES: This study sought to identify disease- and patient-specific characteristics impacting on disease progression and outcomes, and to determine key attributes to be used in developing a conceptual model of multiple myeloma (MM).  METHODS: English language studies reporting MM attributes and their association with disease progression and outcomes were identified from Embase and Medline (2004–2014) and congress abstracts (2012–2014) in a systematic literature review (SLR). A second SLR of treatment guidelines, economic models, health technology reports and studies on the burden of MM was also used. An attribute was defined as a metric or characteristic of MM that plays a potential role in the disease process. These attributes can be explanatory (e.g. patient characteristics) or dependant (e.g. survival). RESULTS: From both SLRs, 95 MM attributes were identified. These were grouped into disease characteristics (e.g. light chains, International Staging System [ISS] stage, bone marrow plasma cells, extramedullary disease, karyotypic abnormalities), genetic factors [e.g. t(4;14), del(13p), del(17p), hypodiploidy, hyperdiploidy], patient characteristics (e.g. age, serum lactate dehydrogenase, gender, Eastern Cooperative Oncology Group performance status, comorbidities), outcomes (e.g. overall survival), quality of life and symptoms (e.g. pain, fatigue, weakness, bone fractures, infection). Attributes were then categorised as explanatory or dependent variables. Age, serum lactate dehydrogenase, light chains and M protein were among the most common explanatory variables in the literature. The most commonly reported dependent variable was overall survival, followed by quality of life. None of the studies presented a comprehensive set of determinants of disease progression and outcomes. CONCLUSIONS: MM is a heterogeneous condition and it is not yet clear which attributes play a key role in determining disease progression and survival. The next step in developing the conceptual model for MM is to ask physicians to validate the potential attributes identified and to clarify the relationships between validated attributes.