Background We have no clear overview of the extent to which health‐care providers involve patients in the decision‐making process during consultations. The Observing Patient Involvement in Decision Making instrument (OPTION) was designed to assess this. Objective To systematically review studies that used the OPTION instrument to observe the extent to which health‐care providers involve patients in decision making across a range of clinical contexts, including different health professions and lengths of consultation. Search strategy We conducted online literature searches in multiple databases (2001–12) and gathered further data through networking. Inclusion criteria (i) OPTION scores as reported outcomes and (ii) health‐care providers and patients as study participants. For analysis, we only included studies using the revised scale. Data extraction Extracted data included: (i) study and participant characteristics and (ii) OPTION outcomes (scores, statistical associations and reported psychometric results). We also assessed the quality of OPTION outcomes reporting. Main results We found 33 eligible studies, 29 of which used the revised scale. Overall, we found low levels of patient‐involving behaviours: in cases where no intervention was used to implement shared decision making (SDM), the mean OPTION score was 23 ± 14 (0–100 scale). When assessed, the variables most consistently associated with higher OPTION scores were interventions to implement SDM (n = 8/9) and duration of consultations (n = 8/15). Conclusions Whatever the clinical context, few health‐care providers consistently attempt to facilitate patient involvement, and even fewer adjust care to patient preferences. However, both SDM interventions and longer consultations could improve this.