BackgroundAlthough evidence shows that poor air quality can harm human health, we have a limited understanding about the behavioural impact of air quality forecasts. Our aim was to understand to what extent air quality warning systems influence protective behaviours in the general public, and to identify the demographic and psychosocial factors associated with adherence and non-adherence to the health advice accompanying these warnings.MethodIn August 2016 literature was systematically reviewed to find studies assessing intended or actual adherence to health advice accompanying air quality warning systems, and encouraging people to reduce exposure to air pollution. Predictors of adherence to the health advice and/or self-reported reasons for adherence or non-adherence were also systematically reviewed. Studies were included only if they involved participants who were using or were aware of these warning systems. Studies investigating only protective behaviours due to subjective perception of bad air quality alone were excluded. The results were narratively synthesised and discussed within the COM-B theoretical framework.ResultsTwenty-one studies were included in the review: seventeen investigated actual adherence; three investigated intended adherence; one assessed both. Actual adherence to the advice to reduce or reschedule outdoor activities during poor air quality episodes ranged from 9.7% to 57% (Median = 31%), whereas adherence to a wider range of protective behaviours (e.g. avoiding busy roads, taking preventative medication) ranged from 17.7% to 98.1% (Median = 46%). Demographic factors did not consistently predict adherence. However, several psychosocial facilitators of adherence were identified. These include knowledge on where to check air quality indices, beliefs that one’s symptoms were due to air pollution, perceived severity of air pollution, and receiving advice from health care professionals. Barriers to adherence included: lack of understanding of the indices, being exposed to health messages that reduced both concern about air pollution and perceived susceptibility, as well as perceived lack of self-efficacy/locus of control, reliance on sensory cues and lack of time.ConclusionWe found frequent suboptimal adherence rates to health advice accompanying air quality alerts. Several psychosocial facilitators and barriers of adherence were identified. To maximise their health effects, health advice needs to target these specific psychosocial factors.Electronic supplementary materialThe online version of this article (10.1186/s12940-017-0307-4) contains supplementary material, which is available to authorized users.